{"slug": "openrouter", "name": "OpenRouter", "category": "ai", "category_name": "AI models and assistants", "description": "Unified API that routes requests to many AI models and providers with one key.", "recommenders": 62, "commenters": 189, "warners": 16, "score": 70.371, "rising": 21.53, "first": "2025-11-02", "latest": "2026-10-31", "why": "Liked for switching between many models and providers through one API key and one bill, and for pay-as-you-go credits that commenters find cheaper than subscriptions.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 62, "neutral": 116, "compare": 11, "warn": 17}, "spark": [0, 1, 2, 2, 3, 1, 6, 7, 9, 12, 11, 8], "listed": true, "loved": ["Cheap metered usage", "Easy model switching"], "caveat": "Unreliable routing", "short_quote": "I switched to just using OpenRouter a year ago because LiteLLM was incredible flaky and kept causing outages", "icon": true, "domain": "openrouter.ai", "github_repo": null, "official_url": "https://openrouter.ai", "pricing": "freemium", "aliases": ["OpenRouter"], "verdict": [{"text": "Liked for switching between many models and providers through one API key and one bill", "cites": [{"id": 794512976, "item_id": 49657061, "url": "https://news.ycombinator.com/item?id=49657061", "by": "embedding-shape", "date": "2026-09-11", "stance": "recommend", "quote": "This is quite literally the point of OpenRouter. A unified interface, so you can easily switch providers without changing a ton of code", "excerpt": "> just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. But that's the intention right? Even the name implies they just send stuff around for you, and if you want to control the routing, you'd lock down providers. I don't see how they could build what they wanted to build, and not have it end up unreliable if you freely round-robin between providers, i", "context": "", "use_case": "unified LLM provider interface", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 789449120, "item_id": 49340570, "url": "https://news.ycombinator.com/item?id=49340570", "by": "PreownedPlaid", "date": "2026-08-17", "stance": "recommend", "quote": "Switching models without opening a new vendor account is the thing I use it for.", "excerpt": "The product people actually pay for is one bill and one API key. Routing is the marketing. Switching models without opening a new vendor account is the thing I use it for.", "context": "", "use_case": "switching models via one API key", "switched_from": "", "story": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter"}, {"id": 781022736, "item_id": 48813921, "url": "https://news.ycombinator.com/item?id=48813921", "by": "KetoManx64", "date": "2026-07-06", "stance": "recommend", "quote": "I use OpenRoutet which lets you switch between providers (Anthropic, ChatGPT, Z-AI) whenever you want.", "excerpt": "Most of the AI companies have OpenAi compatible API's, so you just get a subscription from another provider and change the URL that your LLM Agent Harness uses to talk to the AI. I use OpenRoutet which lets you switch between providers (Anthropic, ChatGPT, Z-AI) whenever you want. Sometimes I'll have two different models from different providers evaluate each other's answers.", "context": "", "use_case": "switching LLM providers", "switched_from": "", "story": "GLM 5.2 and the coming AI margin collapse"}, {"id": 790570098, "item_id": 49410631, "url": "https://news.ycombinator.com/item?id=49410631", "by": "microtonal", "date": "2026-08-23", "stance": "recommend", "quote": "Use OpenRouter or HF Inference Providers in between and you can effortlessly switch between models and providers.", "excerpt": "peace of mind versus using some Chinese cloud These are open weight models (GLM-5.3 soon too). You can run them on the Together AIs or Firework AIs of this world. Use OpenRouter or HF Inference Providers in between and you can effortlessly switch between models and providers. I have been using GLM and Kimi models the last few months mixed with the latest Anthropic models and for my daily work there is barely a differ", "context": "", "use_case": "switching models and providers", "switched_from": "", "story": "GLM-5.3 (open-weight) beat Anthropic/OpenAI models \u2013 for 1/5 the cost"}]}, {"text": ", and for pay-as-you-go credits that commenters find cheaper than subscriptions.", "cites": [{"id": 799989888, "item_id": 49999368, "url": "https://news.ycombinator.com/item?id=49999368", "by": "MisterMunchkin", "date": "2026-10-07", "stance": "recommend", "quote": "I use the free versions on mobile, and openrouter on desktop. It\u2019s way cheaper than paying a subscription", "excerpt": "I use the free versions on mobile, and openrouter on desktop. It\u2019s way cheaper than paying a subscription, and I can always use the latest models.", "context": "", "use_case": "access to latest models", "switched_from": "", "story": "I'm not paying $20 for ChatGPT or Claude because a free local LLM does"}, {"id": 790656816, "item_id": 49416051, "url": "https://news.ycombinator.com/item?id=49416051", "by": "sznio", "date": "2026-08-24", "stance": "recommend", "quote": "I just put $30 on openrouter, switched to Pi, and I finally have a calm mind.", "excerpt": "In my case it's been kind of unhealthy, just constantly waiting for the token limit to reset, always feeling like I'm wasting a resource if I'm not using subscription right now. I just put $30 on openrouter, switched to Pi, and I finally have a calm mind. Since I actually pay per request I want to maximize efficiency rather than utilization", "context": "", "use_case": "pay-per-request model access", "switched_from": "Claude", "story": "Anthropic's best AI model struggles to attract users as cheaper tools thrive"}, {"id": 789676544, "item_id": 49354784, "url": "https://news.ycombinator.com/item?id=49354784", "by": "oceanplexian", "date": "2026-08-18", "stance": "recommend", "quote": "$200 buys a lot of tokens on OpenRouter.", "excerpt": "I use the $200 plan w/ Anthropic and run out of tokens half way through the week and supposedly they are progressively reducing the limits on all their subs even further. At some point I will switch, $200 buys a lot of tokens on OpenRouter.", "context": "pays for the $200 Anthropic plan", "use_case": "cheaper LLM tokens", "switched_from": "", "story": "GLM-5.3 Artificial Analysis Benchmarks"}, {"id": 779068960, "item_id": 48691810, "url": "https://news.ycombinator.com/item?id=48691810", "by": "stanac", "date": "2026-06-26", "stance": "recommend", "quote": "I switched to openrouter and have enough prepaid credits to last me for months with Chinese models.", "excerpt": "Last time I used subscription I hit a usage limit in 20 minutes (gemini). I switched to openrouter and have enough prepaid credits to last me for months with Chinese models. I spent about $30 in last two months.", "context": "", "use_case": "pay-per-use model access", "switched_from": "Gemini", "story": "U.S. government will decide who gets to use GPT-5.6"}]}, {"text": " The most common complaint is unreliable automatic routing that forces users to pin providers", "cites": [{"id": 794501328, "item_id": 49656333, "url": "https://news.ycombinator.com/item?id=49656333", "by": "joshstrange", "date": "2026-09-11", "stance": "warn", "quote": "It\u2019s just incredibly unreliable and you are forced to pin providers", "excerpt": "This squares with my, much much, smaller OpenRouter usage. It\u2019s just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. OpenRouter sells the idea of swapping being commodity providers but it couldn\u2019t be further from the truth. Provider A is often not swappable for B or C (again, as this author found). It can be crazy-making as you sit there thinking \u201cOpe", "context": "", "use_case": "LLM routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 794625216, "item_id": 49664076, "url": "https://news.ycombinator.com/item?id=49664076", "by": "pessimizer", "date": "2026-09-11", "stance": "warn", "quote": "If it doesn't do that, then what is it good for? If I have to permanently pin the one provider who doesn't ignore what I've asked, why shouldn't I just deal with them directly?", "excerpt": "> Parent is complaining about the auto-provider chosing, not that all providers are unreliable. Parent is claiming that the choosing is unreliable. If I rely on a provider to bring me tuna to some spec, but they get it from many different fishermen, it doesn't mean that the tuna doesn't have meet the spec. The complaint is that they're given a bunch of knobs that simply don't work with providers that they could be sw", "context": "", "use_case": "LLM provider routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 795154673, "item_id": 49697167, "url": "https://news.ycombinator.com/item?id=49697167", "by": "Roark66", "date": "2026-09-14", "stance": "warn", "quote": "Openrouter is not ideal, because you don't know who they send your traffic to", "excerpt": "Openrouter is not ideal, because you don't know who they send your traffic to and there are rumours of vendors cheating by providing quantized models. I'm very happy with Deepinfra. Less model coverage, but good prices and quite fast. However I have to caution you about one thing. No one will give you as many input tokens for so little money as Claude Max x5 (maybe x20 too, I use x5). I tend to use 1.1B to 1.4B a wee", "context": "", "use_case": "", "switched_from": "", "story": "Big AI sets out its terms for regulatory capture"}]}, {"text": ", with others reporting higher costs than going direct because of markup and broken caching.", "cites": [{"id": 778711984, "item_id": 48669499, "url": "https://news.ycombinator.com/item?id=48669499", "by": "tacomagick", "date": "2026-06-24", "stance": "warn", "quote": "Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API", "excerpt": "For my case Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API, with 2$ I was able to get around 120M tokens from deepseek easily when Openrouter could only barely do 250k", "context": "", "use_case": "accessing DeepSeek models", "switched_from": "", "story": "GLM-5.2 is the step change for open agents"}, {"id": 786449057, "item_id": 49153066, "url": "https://news.ycombinator.com/item?id=49153066", "by": "thecopy", "date": "2026-08-03", "stance": "warn", "quote": "Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern.", "excerpt": "I agree with parent. OpenRouter might be cheaper list-price, but i have been using 10$ on DS platform since April/May, still have 2$ left. Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern. No idea why.", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "Qwen3.8-Max: A New Bar for Coding and Cowork"}, {"id": 796736177, "item_id": 49796011, "url": "https://news.ycombinator.com/item?id=49796011", "by": "oulu2006", "date": "2026-09-22", "stance": "warn", "quote": "I find it cheaper than paying the openrouter tax", "excerpt": "So everyone has their preferred way of doing it. 1. I go direct to source, i.e. DS platform, I find it cheaper than paying the openrouter tax -- I also switch it up a bit 2. I built a local LLM router, that I update with new profiles that have my preferred provider of the week (lowest token costs/speed) with fallbacks, like mimo --> DS4 etc.. if there is overloading, 3. I use 3 diff harnesses, CC/Codex + Opencode -- ", "context": "", "use_case": "model routing", "switched_from": "", "story": "Elevated errors for multiple models \u2013 Resolved"}]}], "praise": [{"label": "Cheap metered usage", "summary": "Pay-per-use credits last a long time and often cost less than subscriptions or hitting usage limits.", "quotes": [{"id": 803792992, "item_id": 50237062, "url": "https://news.ycombinator.com/item?id=50237062", "by": "schaefer", "date": "2026-10-25", "stance": "recommend", "quote": "At home, I mostly use openrouter.\n\nA few bucks will realistically last you forever.", "excerpt": "> \"What kind of hardware... \" At home, I mostly use openrouter. A few bucks will realistically last you forever. I also bought a DGX Spark at launch. I've learned a ton. but it's not usually my daily driver. I do run overnight tasks on it, like \"read this book and write up notes on each chapter...\" In retrospect, given the dirt cheap prices of apis and or renting hardware, buying hardware for inference was likely a m", "context": "owns a DGX Spark", "use_case": "running LLMs cheaply", "switched_from": "", "story": "OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?"}, {"id": 790656816, "item_id": 49416051, "url": "https://news.ycombinator.com/item?id=49416051", "by": "sznio", "date": "2026-08-24", "stance": "recommend", "quote": "I just put $30 on openrouter, switched to Pi, and I finally have a calm mind.", "excerpt": "In my case it's been kind of unhealthy, just constantly waiting for the token limit to reset, always feeling like I'm wasting a resource if I'm not using subscription right now. I just put $30 on openrouter, switched to Pi, and I finally have a calm mind. Since I actually pay per request I want to maximize efficiency rather than utilization", "context": "", "use_case": "pay-per-request model access", "switched_from": "Claude", "story": "Anthropic's best AI model struggles to attract users as cheaper tools thrive"}, {"id": 783253377, "item_id": 48953336, "url": "https://news.ycombinator.com/item?id=48953336", "by": "KetoManx64", "date": "2026-07-17", "stance": "recommend", "quote": "I use GLM 5.2 through open router", "excerpt": "I use GLM 5.2 through open router, and it's much much less expensive than Anthropic/GPT5.5. I do app development with it and I spend maybe $0.20-$0.50 per task that I work on expanding the app where Claude sonnet will come out to $1-2 for the same work, and Claude Opus $5-10", "context": "app developer", "use_case": "accessing models", "switched_from": "", "story": "Kimi K3 is now #1 in the Front end Code Arena with 1679 pts, surpassing Fable 5"}]}, {"label": "Easy model switching", "summary": "One API lets users swap models and providers without new accounts or code changes.", "quotes": [{"id": 794512976, "item_id": 49657061, "url": "https://news.ycombinator.com/item?id=49657061", "by": "embedding-shape", "date": "2026-09-11", "stance": "recommend", "quote": "This is quite literally the point of OpenRouter. A unified interface, so you can easily switch providers without changing a ton of code", "excerpt": "> just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. But that's the intention right? Even the name implies they just send stuff around for you, and if you want to control the routing, you'd lock down providers. I don't see how they could build what they wanted to build, and not have it end up unreliable if you freely round-robin between providers, i", "context": "", "use_case": "unified LLM provider interface", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 789449120, "item_id": 49340570, "url": "https://news.ycombinator.com/item?id=49340570", "by": "PreownedPlaid", "date": "2026-08-17", "stance": "recommend", "quote": "Switching models without opening a new vendor account is the thing I use it for.", "excerpt": "The product people actually pay for is one bill and one API key. Routing is the marketing. Switching models without opening a new vendor account is the thing I use it for.", "context": "", "use_case": "switching models via one API key", "switched_from": "", "story": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter"}, {"id": 781393280, "item_id": 48837080, "url": "https://news.ycombinator.com/item?id=48837080", "by": "mrinterweb", "date": "2026-07-08", "stance": "recommend", "quote": "I love the model of OpenRouter and being able to switch models at will", "excerpt": "I really don't want harness lock-in. I am trying to decouple myself from Claude Code now. I love the model of OpenRouter and being able to switch models at will let's your harness focus on your personal tooling and you can easily switch to the flavor of the month LLM with a single slash command instead of rewiring your entire workflow to use a harness to use a model. I like Cerabras, but I really wish they would make", "context": "", "use_case": "switching LLMs at will", "switched_from": "Claude Code", "story": "SWE-1.7 Reach Near GPT 5.5 and Opus Intelligence"}]}, {"label": "Single balance, spend caps", "summary": "One prepaid balance across providers limits how much a runaway agent can spend.", "quotes": [{"id": 789301536, "item_id": 49331346, "url": "https://news.ycombinator.com/item?id=49331346", "by": "Macha", "date": "2026-08-17", "stance": "recommend", "quote": "the big reason I use OpenRouter rather than the providers directly is the ability to cap spending", "excerpt": "For me the big reason I use OpenRouter rather than the providers directly is the ability to cap spending. With OpenRouter I can top up by $50, change models on a whim and if some agent loop goes out of control and generates a ton of tokens, the most I can be out is $50, while using the providers directly requires me to manage accounts there and not all support top up based spending.", "context": "", "use_case": "capped-spend model access", "switched_from": "", "story": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter"}, {"id": 794569648, "item_id": 49660603, "url": "https://news.ycombinator.com/item?id=49660603", "by": "arjie", "date": "2026-09-11", "stance": "recommend", "quote": "OpenRouter has the advantage that I can maintain a single balance across model providers so I use it for quick iteration", "excerpt": "OpenRouter has the advantage that I can maintain a single balance across model providers so I use it for quick iteration across and then write it into my own router (required anyway since I have many on-prem models running). The portable balance is actually very useful and my credit card is in one place. If I could pay per request without maintaining a balance or credit card out of a single wallet (using crypto or so", "context": "runs many on-prem models", "use_case": "quick iteration across model providers", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 794577856, "item_id": 49661116, "url": "https://news.ycombinator.com/item?id=49661116", "by": "habosa", "date": "2026-09-11", "stance": "recommend", "quote": "We're very happy with OpenRouter in production", "excerpt": "We're very happy with OpenRouter in production, although we use it in a pretty limited fashion. We have only ~10 providers allowlisted (based on their location and ZDR policies) and we use only a few models, often frontier models. Has been very reliable for us in production so far. It's generally been great for us to be able to have a single place to manage spend, policies, view logs, etc. Also really nice in code to", "context": "production user with allowlisted providers", "use_case": "LLM routing in production", "switched_from": "", "story": "So you want to use OpenRouter?"}]}, {"label": "Zero data retention", "summary": "ZDR options let users run models, including Chinese ones, without retention or training.", "quotes": [{"id": 800217184, "item_id": 50013574, "url": "https://news.ycombinator.com/item?id=50013574", "by": "patwolf", "date": "2026-10-08", "stance": "recommend", "quote": "One of the reasons I use OpenRouter is because they offer zero data retention.", "excerpt": "One of the reasons I use OpenRouter is because they offer zero data retention. As far as I can tell, DeepSeek's own API doesn't support ZDR.", "context": "", "use_case": "LLM API access with ZDR", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 796262420, "item_id": 49766401, "url": "https://news.ycombinator.com/item?id=49766401", "by": "Frannky", "date": "2026-09-19", "stance": "recommend", "quote": "the ability to use Chinese models on international providers with no training and ZDR behind OpenRouter are very powerful", "excerpt": "I switched to oh-my-py and never looked back. I guess having an incentive to maximize (expensive) token usage, and LLMs degrading with higher context, isn't aligned with a user's best interest. Oh-my-py unlocks max capabilities at the smartest token usage, and you can change AI company and providers while still getting things done. Roles, advisor, and the ability to use Chinese models on international providers with ", "context": "", "use_case": "model routing", "switched_from": "", "story": "Claude Code now reads AGENTS.md if there is no Claude.md"}, {"id": 794577856, "item_id": 49661116, "url": "https://news.ycombinator.com/item?id=49661116", "by": "habosa", "date": "2026-09-11", "stance": "recommend", "quote": "We're very happy with OpenRouter in production", "excerpt": "We're very happy with OpenRouter in production, although we use it in a pretty limited fashion. We have only ~10 providers allowlisted (based on their location and ZDR policies) and we use only a few models, often frontier models. Has been very reliable for us in production so far. It's generally been great for us to be able to have a single place to manage spend, policies, view logs, etc. Also really nice in code to", "context": "production user with allowlisted providers", "use_case": "LLM routing in production", "switched_from": "", "story": "So you want to use OpenRouter?"}]}, {"label": "Console and telemetry", "summary": "The console, logs and telemetry make it easy to track usage and spot key misuse.", "quotes": [{"id": 781069713, "item_id": 48816857, "url": "https://news.ycombinator.com/item?id=48816857", "by": "pixlmint", "date": "2026-07-07", "stance": "recommend", "quote": "Openrouter has very good telemetry (which is partly why I went with them)", "excerpt": "I have OpenWebui hosted on my homelab, but you can also just have it live on your machine in a docker container. I honestly just embrace the iffy feeling. Openrouter has very good telemetry (which is partly why I went with them) and it'd be pretty easy to notice when someone other than myself uses my keys. For the little agentic coding I do I like to use OpenCode, and if I need to ask a question in my editor I use Co", "context": "homelab user", "use_case": "LLM API access", "switched_from": "", "story": "GLM 5.2 and the coming AI margin collapse"}, {"id": 794577856, "item_id": 49661116, "url": "https://news.ycombinator.com/item?id=49661116", "by": "habosa", "date": "2026-09-11", "stance": "recommend", "quote": "We're very happy with OpenRouter in production", "excerpt": "We're very happy with OpenRouter in production, although we use it in a pretty limited fashion. We have only ~10 providers allowlisted (based on their location and ZDR policies) and we use only a few models, often frontier models. Has been very reliable for us in production so far. It's generally been great for us to be able to have a single place to manage spend, policies, view logs, etc. Also really nice in code to", "context": "production user with allowlisted providers", "use_case": "LLM routing in production", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 783385047, "item_id": 48961565, "url": "https://news.ycombinator.com/item?id=48961565", "by": "ignoramous", "date": "2026-07-18", "stance": "recommend", "quote": "I always put my coding subscriptions through \"AI gateways\" (Cloudflare & OpenRouter are free) which help track token use.", "excerpt": "> tried Kimi K3 on a task I've done with every other model I use regularly and found it chewed a lot longer on the problem and ate up almost the entirety of a 5 hour usage limit on their $19 plan ArtificialAnalysis puts Kimi K3 just below DeepSeek v4 & GLM 5.2 in token use per task, which is about 2x to 3x more tokens than Grok 4.5: https://x.com/ArtificialAnlys/status/2077832879187620192 > Subscription usage limits ", "context": "", "use_case": "track token use", "switched_from": "", "story": "The Kimi K3 Moment"}]}, {"label": "Free models", "summary": "Free models are offered from time to time.", "quotes": [{"id": 803529921, "item_id": 50220620, "url": "https://news.ycombinator.com/item?id=50220620", "by": "alwinaugustin", "date": "2026-10-24", "stance": "recommend", "quote": "Opencode or Openrouter allows free models time to time.", "excerpt": "I use free models only these days. Opencode or Openrouter allows free models time to time.", "context": "", "use_case": "free models", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 795889121, "item_id": 49743070, "url": "https://news.ycombinator.com/item?id=49743070", "by": "alexjplant", "date": "2026-09-17", "stance": "recommend", "quote": "the a la carte pricing via OpenRouter makes this a non-issue", "excerpt": "Just use GLM-5.3 Flash via OpenRouter. It's dirt cheap especially relative to how capable it is. While the Z.ai coding plan was a decent deal in the past I always ran into limiting with it and since I use it intermittently for personal projects my usage wasn't always enough to make the math work - the a la carte pricing via OpenRouter makes this a non-issue. There's also a new free stealth model available that's more", "context": "", "use_case": "pay-per-use model access", "switched_from": "", "story": "How GLM built its own inference infrastructure"}]}], "caveats": [{"label": "Unreliable routing", "summary": "Automatic routing across providers gives inconsistent quality and errors, so users pin providers.", "quotes": [{"id": 794501328, "item_id": 49656333, "url": "https://news.ycombinator.com/item?id=49656333", "by": "joshstrange", "date": "2026-09-11", "stance": "warn", "quote": "It\u2019s just incredibly unreliable and you are forced to pin providers", "excerpt": "This squares with my, much much, smaller OpenRouter usage. It\u2019s just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. OpenRouter sells the idea of swapping being commodity providers but it couldn\u2019t be further from the truth. Provider A is often not swappable for B or C (again, as this author found). It can be crazy-making as you sit there thinking \u201cOpe", "context": "", "use_case": "LLM routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 794625216, "item_id": 49664076, "url": "https://news.ycombinator.com/item?id=49664076", "by": "pessimizer", "date": "2026-09-11", "stance": "warn", "quote": "If it doesn't do that, then what is it good for? If I have to permanently pin the one provider who doesn't ignore what I've asked, why shouldn't I just deal with them directly?", "excerpt": "> Parent is complaining about the auto-provider chosing, not that all providers are unreliable. Parent is claiming that the choosing is unreliable. If I rely on a provider to bring me tuna to some spec, but they get it from many different fishermen, it doesn't mean that the tuna doesn't have meet the spec. The complaint is that they're given a bunch of knobs that simply don't work with providers that they could be sw", "context": "", "use_case": "LLM provider routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 798841872, "item_id": 49927617, "url": "https://news.ycombinator.com/item?id=49927617", "by": "adchurch", "date": "2026-10-01", "stance": "recommend", "quote": "OpenRouter does make it much easier to get started (at the cost of that variance potentially affecting quality).", "excerpt": "For our hosted version we carefully select which providers we use (we don't use OpenRouter). For the self-hosted version though, OpenRouter does make it much easier to get started (at the cost of that variance potentially affecting quality). Yes for both! We have some logic to put providers on cooldowns and/or deprioritize them.", "context": "built a model routing product", "use_case": "model provider routing for self-hosted", "switched_from": "", "story": "Show HN: Open-source model routing for coding agents at Astra-level performance"}]}, {"label": "Hidden costs", "summary": "Some report markup, broken caching or expensive provider picks making it costlier than direct APIs.", "quotes": [{"id": 778711984, "item_id": 48669499, "url": "https://news.ycombinator.com/item?id=48669499", "by": "tacomagick", "date": "2026-06-24", "stance": "warn", "quote": "Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API", "excerpt": "For my case Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API, with 2$ I was able to get around 120M tokens from deepseek easily when Openrouter could only barely do 250k", "context": "", "use_case": "accessing DeepSeek models", "switched_from": "", "story": "GLM-5.2 is the step change for open agents"}, {"id": 786449057, "item_id": 49153066, "url": "https://news.ycombinator.com/item?id=49153066", "by": "thecopy", "date": "2026-08-03", "stance": "warn", "quote": "Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern.", "excerpt": "I agree with parent. OpenRouter might be cheaper list-price, but i have been using 10$ on DS platform since April/May, still have 2$ left. Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern. No idea why.", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "Qwen3.8-Max: A New Bar for Coding and Cowork"}, {"id": 779184865, "item_id": 48699054, "url": "https://news.ycombinator.com/item?id=48699054", "by": "peheje", "date": "2026-06-27", "stance": "warn", "quote": "OpenRouter sometimes chooses a very expensive provider.", "excerpt": "OpenRouter sometimes chooses a very expensive provider. Try the floor slug or choose directly the provider. I moved to just putting 5 dollars directly on deepseek instead of going through OR.", "context": "", "use_case": "routing model API requests", "switched_from": "", "story": "DeepSeek open-sources inference optimizations with 60\u201385% faster generation [pdf]"}]}, {"label": "Billing surprises", "summary": "One user saw $50 consumed in under an hour while their agent reported about $1.", "quotes": [{"id": 783469792, "item_id": 48966862, "url": "https://news.ycombinator.com/item?id=48966862", "by": "onlyrealcuzzo", "date": "2026-07-19", "stance": "warn", "quote": "it burned through $50 in under an hour on pretty trivial stuff.", "excerpt": "Who do you buy DeepSeek from? I bought it through OpenRouter and used it with Pi agent. The model was good, but there appeared to be a pricing glitch or something, because it burned through $50 in under an hour on pretty trivial stuff. Pi agent claimed it only used like $1. OpenRouter claimed differently and said I used all $50.", "context": "", "use_case": "buying DeepSeek access", "switched_from": "", "story": "Qwen3.8 is launching and going open-weight soon"}]}, {"label": "Trust and privacy", "summary": "Some distrust it over unclear routing destinations and a change to data retention settings.", "quotes": [{"id": 801876656, "item_id": 50117291, "url": "https://news.ycombinator.com/item?id=50117291", "by": "paweladamczuk", "date": "2026-10-16", "stance": "warn", "quote": "My trust in Openrouter took a massive hit today.", "excerpt": "I had Zero Data Retention enabled for all models I use through Openrouter. I was sure I was safe from my data being saved - if I tried to use a model with no ZDR guarantees, Openrouter would just return an error code through the API. This is how I had my guardrails configured. On October 16 I used the Openrouter site when logged in and I noticed a banner about new Safety Retention features for Google models. I though", "context": "uses it for clients", "use_case": "accessing models with zero data retention", "switched_from": "", "story": "Google models no longer Zero Data Retention on Openrouter"}, {"id": 791679777, "item_id": 49479986, "url": "https://news.ycombinator.com/item?id=49479986", "by": "alexnewman", "date": "2026-08-28", "stance": "warn", "quote": "Open router is not to be trusted.", "excerpt": "Open router is not to be trusted. It\u2019s not open. They\u2019ve misled people on logs. They are not reliable. I used Trusted router as it is open source", "context": "", "use_case": "LLM API routing", "switched_from": "", "story": "OpenRouter Is Having Issues"}, {"id": 795154673, "item_id": 49697167, "url": "https://news.ycombinator.com/item?id=49697167", "by": "Roark66", "date": "2026-09-14", "stance": "warn", "quote": "Openrouter is not ideal, because you don't know who they send your traffic to", "excerpt": "Openrouter is not ideal, because you don't know who they send your traffic to and there are rumours of vendors cheating by providing quantized models. I'm very happy with Deepinfra. Less model coverage, but good prices and quite fast. However I have to caution you about one thing. No one will give you as many input tokens for so little money as Claude Max x5 (maybe x20 too, I use x5). I tend to use 1.1B to 1.4B a wee", "context": "", "use_case": "", "switched_from": "", "story": "Big AI sets out its terms for regulatory capture"}]}, {"label": "Hard to debug", "summary": "Empty responses returned with success codes and provider issues can be hard to diagnose.", "quotes": [{"id": 794523456, "item_id": 49657716, "url": "https://news.ycombinator.com/item?id=49657716", "by": "neya", "date": "2026-09-11", "stance": "warn", "quote": "The best part about OpenRouter is 200 OK is probably hardcoded into their responses.", "excerpt": "The best part about OpenRouter is 200 OK is probably hardcoded into their responses. I used to get content: \"\" all the time and I used to triple check my code to see if I was doing something wrong until I realized most AI providers in general have vibe coded their infrastructure as well and it is just a futile attempt to even fight it.", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 789202512, "item_id": 49325157, "url": "https://news.ycombinator.com/item?id=49325157", "by": "truetraveller", "date": "2026-08-16", "stance": "warn", "quote": "I had consistent problems (5 months back) with OpenRouter and other providers. And very hard to debug.", "excerpt": "I had consistent problems (5 months back) with OpenRouter and other providers. And very hard to debug. I used the provider directly (Google) and worked perfectly.", "context": "", "use_case": "LLM API routing", "switched_from": "", "story": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter"}]}], "who": [{"by": "schaefer", "context": "owns a DGX Spark", "url": "https://news.ycombinator.com/item?id=50237062", "date": "2026-10-25"}, {"by": "adchurch", "context": "built a model routing product", "url": "https://news.ycombinator.com/item?id=49927617", "date": "2026-10-01"}, {"by": "abalashov", "context": "individual contributor", "url": "https://news.ycombinator.com/item?id=49815221", "date": "2026-09-23"}, {"by": "glub", "context": "developer/founder", "url": "https://news.ycombinator.com/item?id=49668549", "date": "2026-09-12"}, {"by": "habosa", "context": "production user with allowlisted providers", "url": "https://news.ycombinator.com/item?id=49661116", "date": "2026-09-11"}, {"by": "arjie", "context": "runs many on-prem models", "url": "https://news.ycombinator.com/item?id=49660603", "date": "2026-09-11"}, {"by": "rolfus", "context": "built a running tracker app", "url": "https://news.ycombinator.com/item?id=49658044", "date": "2026-09-11"}, {"by": "andai", "context": "uses a custom harness", "url": "https://news.ycombinator.com/item?id=49368928", "date": "2026-08-19"}, {"by": "nikcub", "context": "long time user", "url": "https://news.ycombinator.com/item?id=49368013", "date": "2026-08-19"}, {"by": "drob518", "context": "developer", "url": "https://news.ycombinator.com/item?id=49365293", "date": "2026-08-19"}, {"by": "oceanplexian", "context": "pays for the $200 Anthropic plan", "url": "https://news.ycombinator.com/item?id=49354784", "date": "2026-08-18"}, {"by": "everforward", "context": "human-in-the-loop dev process", "url": "https://news.ycombinator.com/item?id=49331937", "date": "2026-08-17"}], "tags": ["for developers", "for agentic coding", "for trying new models", "for production apps", "for budget-conscious users"], "use_cases": ["model routing", "model access", "llm api access", "llm api routing", "accessing models", "model api access"], "compared": [{"slug": "deepseek", "name": "DeepSeek", "co": 32, "switched_from": 0, "switched_to": 1}, {"slug": "claude", "name": "Claude", "co": 21, "switched_from": 4, "switched_to": 0}, {"slug": "opencode", "name": "OpenCode", "co": 27, "switched_from": 0, "switched_to": 1}, {"slug": "claude-code", "name": "Claude Code", "co": 24, "switched_from": 1, "switched_to": 0}, {"slug": "pi", "name": "Pi", "co": 22, "switched_from": 0, "switched_to": 0}, {"slug": "deepseek-v4-flash", "name": "DeepSeek V4 Flash", "co": 16, "switched_from": 0, "switched_to": 1}, {"slug": "glm", "name": "GLM", "co": 16, "switched_from": 0, "switched_to": 0}, {"slug": "kimi", "name": "Kimi", "co": 14, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49621546", "title": "So you want to use OpenRouter?", "mentions": 8}, {"url": "https://news.ycombinator.com/item?id=50000488", "title": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=49364559", "title": "OpenRouter Is Joining Stripe", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=48966120", "title": "Qwen3.8 is launching and going open-weight soon", "mentions": 4}, {"url": "https://news.ycombinator.com/item?id=49323381", "title": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=48999291", "title": "Kimi K3 Is Competitive with Fable; Kimi K3 and Fable Is SoTA", "mentions": 3}], "sentiment_by_month": {"2025-11": {"neutral": 3}, "2025-12": {"neutral": 5, "warn": 1, "recommend": 1}, "2026-01": {"recommend": 2, "neutral": 5}, "2026-02": {"recommend": 2}, "2026-03": {"neutral": 4, "recommend": 3}, "2026-04": {"neutral": 6, "compare": 1, "recommend": 1}, "2026-05": {"neutral": 7, "recommend": 6, "compare": 1}, "2026-06": {"neutral": 16, "warn": 5, "recommend": 7}, "2026-07": {"recommend": 9, "compare": 4, "neutral": 19, "warn": 2}, "2026-08": {"warn": 3, "compare": 3, "neutral": 14, "recommend": 12}, "2026-09": {"neutral": 18, "compare": 2, "recommend": 11, "warn": 5}, "2026-10": {"recommend": 8, "neutral": 19, "warn": 1}}, "quotes": [{"id": 805034641, "item_id": 50314665, "url": "https://news.ycombinator.com/item?id=50314665", "by": "mgrandl", "date": "2026-10-31", "stance": "neutral", "quote": "i switched to span-02 on openrouter", "excerpt": "Just fyi, i switched to span-02 on openrouter as well since it\u2019s 10x cheaper than mammoth and it\u2019s performing just as well for me, if not better because it\u2019s noticably faster! So thanks for the recommendation.", "context": "", "use_case": "model API access", "switched_from": "", "story": "Microsoft Decision-1: 9B Qwen Model Launch Explained"}, {"id": 804634401, "item_id": 50289650, "url": "https://news.ycombinator.com/item?id=50289650", "by": "dichiara19", "date": "2026-10-29", "stance": "neutral", "quote": "I used it via OpenRouter on a Raspberry Pi", "excerpt": "I want to talk about my experience with Argon, noting that I used it via OpenRouter on a Raspberry Pi, spending around $15 in credit. My experience is that Google is aiming to create a generalist model \u2013 but a truly generalist one! In fact, I find it particularly accurate in general terms and even in long conversations, maintaining excellent consistency and a wealth of information in response to complex questions. I\u2019", "context": "", "use_case": "accessing models", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}, {"id": 804621057, "item_id": 50288816, "url": "https://news.ycombinator.com/item?id=50288816", "by": "sgc", "date": "2026-10-29", "stance": "neutral", "quote": "I have specific tough pages I use on openrouter to test every new model I am interested in.", "excerpt": "For ocr, you have to practice, practice, practice with your prompts on small sample sizes before one-shotting a book. I create a json schema for the output that describes the structure in detail (using ai from my prose descriptions of structure of course). I have specific tough pages I use on openrouter to test every new model I am interested in. Gemini has and does absolutely dominate for this. But flash can be bett", "context": "", "use_case": "testing new models", "switched_from": "", "story": "Gemini 4 Carbon is the new #1 on the Artificial Analysis Intelligence Index v5"}, {"id": 804556945, "item_id": 50284809, "url": "https://news.ycombinator.com/item?id=50284809", "by": "killingtime74", "date": "2026-10-29", "stance": "neutral", "quote": "I use openrouter models as well as the local ones", "excerpt": "https://github.com/DevEmperor/DictateKeyboard. I forked it to add more models (with Claude). I use openrouter models as well as the local ones", "context": "", "use_case": "model access for dictation", "switched_from": "", "story": "Nvidia Nemotron 3.5 Transcribe: open ASR for 40 languages"}, {"id": 804192354, "item_id": 50262022, "url": "https://news.ycombinator.com/item?id=50262022", "by": "plqbfbv", "date": "2026-10-27", "stance": "neutral", "quote": "or OpenRouter if you prefer to change models", "excerpt": "I've benchmarked quite a few models on my 9070XT, and I basically gave up hope of using models locally: https://claude.ai/artifact/77pHgucqQNsmic8ZEnqdBA I asked Claude to come up with 7 problems to solve that were not part of any benchmarks. Besides speed, I asked it to put together a scoring that would reflect models that perform well on the 7 problems and in a reasonable time-frame (RWS, the definition is in the a", "context": "", "use_case": "switching between models", "switched_from": "", "story": "New iMac with M6"}, {"id": 803825186, "item_id": 50239074, "url": "https://news.ycombinator.com/item?id=50239074", "by": "bitexploder", "date": "2026-10-26", "stance": "recommend", "quote": "I strongly encourage people to go use this model on OpenRouter and try it out for tasks.", "excerpt": "I mean, I just have 4xV100S GPU and a Macbook Pro w/48GB of RAM and I find it kind of amazing. My server can run Qwen Flash Next at 500 t/s aggregate (about 200 t/s single user). It's pretty great. I use it for local coding and all my worker agents. It is about as reliable as Sonnet 5 for agentic coding. I make 3D models, keep the home network and server updated, fully built ansible files. You do NOT need 20K in hard", "context": "", "use_case": "trying models", "switched_from": "", "story": "IBM CEO says open weight models are now good and cheap on-prem"}, {"id": 803792992, "item_id": 50237062, "url": "https://news.ycombinator.com/item?id=50237062", "by": "schaefer", "date": "2026-10-25", "stance": "recommend", "quote": "At home, I mostly use openrouter.\n\nA few bucks will realistically last you forever.", "excerpt": "> \"What kind of hardware... \" At home, I mostly use openrouter. A few bucks will realistically last you forever. I also bought a DGX Spark at launch. I've learned a ton. but it's not usually my daily driver. I do run overnight tasks on it, like \"read this book and write up notes on each chapter...\" In retrospect, given the dirt cheap prices of apis and or renting hardware, buying hardware for inference was likely a m", "context": "owns a DGX Spark", "use_case": "running LLMs cheaply", "switched_from": "", "story": "OpenAI and Anthropic cut their LLM prices by up to 80% within a few months. Why?"}, {"id": 803538658, "item_id": 50221166, "url": "https://news.ycombinator.com/item?id=50221166", "by": "kelvinjps10", "date": "2026-10-24", "stance": "neutral", "quote": "a few api credits in openrouter", "excerpt": "20$ chatgpt plus sub with pi harness and a few api credits in openrouter.sometiemd I use sol for planing and glm flash for implementation bu", "context": "", "use_case": "API model access", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 803529921, "item_id": 50220620, "url": "https://news.ycombinator.com/item?id=50220620", "by": "alwinaugustin", "date": "2026-10-24", "stance": "recommend", "quote": "Opencode or Openrouter allows free models time to time.", "excerpt": "I use free models only these days. Opencode or Openrouter allows free models time to time.", "context": "", "use_case": "free models", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 803099456, "item_id": 50193716, "url": "https://news.ycombinator.com/item?id=50193716", "by": "luisfominaya", "date": "2026-10-22", "stance": "neutral", "quote": "You can see the model card on OpenRouter since I used them as the inference provider here", "excerpt": "You can see the model card on OpenRouter since I used them as the inference provider here: https://openrouter.ai/openai/gpt-5.6-sol-pro It's basically setting the reasoning mode to 'pro' on OpenAI's API.", "context": "", "use_case": "inference provider", "switched_from": "", "story": "Using LLM for confidence scoring is a bad idea"}, {"id": 802555376, "item_id": 50159711, "url": "https://news.ycombinator.com/item?id=50159711", "by": "simonw", "date": "2026-10-20", "stance": "neutral", "quote": "It only tracks companies that have decided to use OpenRouter to help them switch models", "excerpt": "There are a few problems with using OpenRouter's public statistics to make statements about the wider field: - It only tracks companies that have decided to use OpenRouter to help them switch models - which biases towards companies that like open weights. Anyone using OpenAI or Anthropic directly is much less likely to be represented. - OpenRouter don't share the number of unique users represented in those numbers. I", "context": "", "use_case": "model routing", "switched_from": "", "story": "Token-maxxing is dead, OpenRouter's token usage plateaus"}, {"id": 802516451, "item_id": 50157278, "url": "https://news.ycombinator.com/item?id=50157278", "by": "sinuhe69", "date": "2026-10-19", "stance": "neutral", "quote": "On Openrouter at least", "excerpt": "On Openrouter at least, all Mino models are painfully slow. TTFT is many seconds and the models spend a lot of time on its CoT, making effective output per second many times slower than DS Flash. On Flash 4.1, I constantly get ~180 tok/s and the real work got done fast, too. So even when the Mimo 2.6 are smart models, I use them much less recently. Not to mention their costs are also higher than DS. For me, the DeepS", "context": "", "use_case": "accessing models", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802361169, "item_id": 50147573, "url": "https://news.ycombinator.com/item?id=50147573", "by": "slyall", "date": "2026-10-19", "stance": "neutral", "quote": "Last month I used $0.90 of openrouter.", "excerpt": "Personally at home I use the free github copilot and then openrouter when that runs out. Last month I used $0.90 of openrouter. Zero so far this month. When my bill starts hitting double digits I'll consider options. Different matter at work but even there we are paying for API costs and have a monthly allowance (starts at $200/month but increases fairly easy to get)", "context": "", "use_case": "fallback LLM API access", "switched_from": "", "story": "If you don't run your own LLM proxy, how do you know whats going on?"}, {"id": 802354352, "item_id": 50147147, "url": "https://news.ycombinator.com/item?id=50147147", "by": "impure", "date": "2026-10-19", "stance": "recommend", "quote": "OpenRouter does all this for you. It's pretty nice.", "excerpt": "OpenRouter does all this for you. It's pretty nice. Although for apps I use another server to lock down the requests even more and prevent the client from seeing my API key.", "context": "", "use_case": "LLM proxy and observability", "switched_from": "", "story": "If you don't run your own LLM proxy, how do you know whats going on?"}, {"id": 802082053, "item_id": 50130128, "url": "https://news.ycombinator.com/item?id=50130128", "by": "clx75", "date": "2026-10-17", "stance": "neutral", "quote": "I still had 95 USD on my OpenRouter account", "excerpt": "Probably a week ago someone posted a thread with the title \"Why isn't the industry freaking out about DeepSeek 4.1 Flash?\" I have been intrigued about Chinese models for a long time, but never tried anything except an early Qwen. My daily driver was Luna, with occasional elevation to Terra or Sol - all on 5.6 as I found the output of 6 jarringly incomprehensible. I still had 95 USD on my OpenRouter account, so I thou", "context": "", "use_case": "model API access", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 801876656, "item_id": 50117291, "url": "https://news.ycombinator.com/item?id=50117291", "by": "paweladamczuk", "date": "2026-10-16", "stance": "warn", "quote": "My trust in Openrouter took a massive hit today.", "excerpt": "I had Zero Data Retention enabled for all models I use through Openrouter. I was sure I was safe from my data being saved - if I tried to use a model with no ZDR guarantees, Openrouter would just return an error code through the API. This is how I had my guardrails configured. On October 16 I used the Openrouter site when logged in and I noticed a banner about new Safety Retention features for Google models. I though", "context": "uses it for clients", "use_case": "accessing models with zero data retention", "switched_from": "", "story": "Google models no longer Zero Data Retention on Openrouter"}, {"id": 801801473, "item_id": 50112592, "url": "https://news.ycombinator.com/item?id=50112592", "by": "enjoyyourlife", "date": "2026-10-16", "stance": "neutral", "quote": "I love using it free on Nvidia platform and OpenRouter", "excerpt": "It's only good because it's free. I love using it free on Nvidia platform and OpenRouter but otherwise Nemotron hasn't been good so far. It's nice that they are still working on it but this would have to be a major improvement to justify paying for it", "context": "", "use_case": "accessing models for free", "switched_from": "", "story": "Nvidia Nemotron 3.5 Bolt 74B-A9B"}, {"id": 801251136, "item_id": 50078196, "url": "https://news.ycombinator.com/item?id=50078196", "by": "karahime", "date": "2026-10-13", "stance": "neutral", "quote": "requests that require ZDR will not route to Google authored models", "excerpt": "It's the Vertex endpoint for Gemini models that's changing to be reclassified by OR as not ZDR, as I understand it. The email from them says > Upcoming change to OpenRouter\u2019s ZDR designation for Google authored models > We\u2019re writing to let you know about an upcoming change to how OpenRouter designates Google authored models for Zero Data Retention (ZDR) routing. > Effective October 18, 2026, OpenRouter will no longe", "context": "", "use_case": "routing LLM requests with ZDR", "switched_from": "", "story": "Gemini 4 Argon"}, {"id": 800742610, "item_id": 50046413, "url": "https://news.ycombinator.com/item?id=50046413", "by": "chaosmanage", "date": "2026-10-11", "stance": "neutral", "quote": "It currently uses openrouter using your own key.", "excerpt": "Author here. First of all, thank you to the Harmonic app by Simon Halvdansson (https://github.com/SimonHalvdansson/Harmonic-HN). I know many (most?) of you prefer to use the no frills layout of the hacker news wesite and I totally respect that. But I really enjoy using the Harmonic app on my android phone which has a much more modren layout and is visually appealing and easier to use for me. I disliked using the plai", "context": "", "use_case": "AI summarization backend", "switched_from": "", "story": "Show HN: HarmoNova: A modern, customizable Hacker News reader for Chrome"}, {"id": 800387153, "item_id": 50024197, "url": "https://news.ycombinator.com/item?id=50024197", "by": "sixdimensional", "date": "2026-10-09", "stance": "neutral", "quote": "I can route across OpenRouter, native AI providers directly", "excerpt": "One thing in particular is bifrost[1] as an AI router, and it has worked really well and feels pretty solid. I can route across OpenRouter, native AI providers directly, and even route work to my local AI inference server through tailscale. It makes it very easy to work across providers and platforms. [1] https://github.com/maximhq/bifrost", "context": "", "use_case": "AI provider routing", "switched_from": "", "story": "Ask HN: What do you run on a $5 VPS that's worth keeping online 24/7?"}, {"id": 800229266, "item_id": 50014329, "url": "https://news.ycombinator.com/item?id=50014329", "by": "carsoon", "date": "2026-10-09", "stance": "neutral", "quote": "if i use opus 5.5 in api through openrouter and pi agent harness I will easily burn 50-100$ a day", "excerpt": "Yea if i use opus 5.5 in api through openrouter and pi agent harness I will easily burn 50-100$ a day (and with fable 5.1 i could burn 200$ easily). Whereas i have now been using a claude code subscription for 2 weeks using 5.5 at all times and have never hit a limit. I often run 6+ agent sessions at once. I do think its important long term to not be reliant on these companies as you don't have control over the syste", "context": "", "use_case": "API access to models", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800217184, "item_id": 50013574, "url": "https://news.ycombinator.com/item?id=50013574", "by": "patwolf", "date": "2026-10-08", "stance": "recommend", "quote": "One of the reasons I use OpenRouter is because they offer zero data retention.", "excerpt": "One of the reasons I use OpenRouter is because they offer zero data retention. As far as I can tell, DeepSeek's own API doesn't support ZDR.", "context": "", "use_case": "LLM API access with ZDR", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800205409, "item_id": 50012838, "url": "https://news.ycombinator.com/item?id=50012838", "by": "georgel", "date": "2026-10-08", "stance": "neutral", "quote": "I loaded $100 back in July while using v4-flash", "excerpt": "I am curious how you managed to spend that much on Deepseek via OpenRouter. I loaded $100 back in July while using v4-flash or whatever the cheap good model was at the time, and have upgraded as the new ones came out from Deepseek. I still have $16 and some of that spend also goes towards the AI usage from my customers (the context they need to load in is quite large too). And I am using the Claude Code harness with ", "context": "", "use_case": "API access to DeepSeek", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800200080, "item_id": 50012505, "url": "https://news.ycombinator.com/item?id=50012505", "by": "NortySpock", "date": "2026-10-08", "stance": "neutral", "quote": "I use OpenRouter, I know switching is easy", "excerpt": "Agree that people leaving the big two companies is going to be hard to keep a pulse on prior to IPO. Anthropic and OpenAi are in the news, so they get the press and people go and try out their product. Large enterprise businesses are going to make larger, longer-term contracts with them and are only going to pivot if they think switching costs are easy or if they think the provider won't deliver. The other inference ", "context": "", "use_case": "switching between model providers", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 799989888, "item_id": 49999368, "url": "https://news.ycombinator.com/item?id=49999368", "by": "MisterMunchkin", "date": "2026-10-07", "stance": "recommend", "quote": "I use the free versions on mobile, and openrouter on desktop. It\u2019s way cheaper than paying a subscription", "excerpt": "I use the free versions on mobile, and openrouter on desktop. It\u2019s way cheaper than paying a subscription, and I can always use the latest models.", "context": "", "use_case": "access to latest models", "switched_from": "", "story": "I'm not paying $20 for ChatGPT or Claude because a free local LLM does"}, {"id": 799419217, "item_id": 49963701, "url": "https://news.ycombinator.com/item?id=49963701", "by": "Oras", "date": "2026-10-05", "stance": "recommend", "quote": "always used OpenRouter which does support web search", "excerpt": "Weird choice by CloudFlare, would been great if they have shared why it was created. I use CloudFlare developer platform and quite happy with tools, but I didn\u2019t use the gateway API and always used OpenRouter which does support web search. I can see it useful for those who didn\u2019t do any integrations or like to keep logs at one place, but did customers actually ask for this?", "context": "", "use_case": "LLM API gateway", "switched_from": "", "story": "Web Search API"}, {"id": 799196145, "item_id": 49949759, "url": "https://news.ycombinator.com/item?id=49949759", "by": "tkgally", "date": "2026-10-04", "stance": "neutral", "quote": "I tell Claude to keep to hard daily limits on its OpenRouter spending", "excerpt": "> In an ideal world, our agents could help with this. Not exactly the sort of case Simon has in mind, but I tell Claude to keep to hard daily limits on its OpenRouter spending for two long-running projects [1, 2]. A Routine for each project fires ever few hours, and Claude decides itself what to do in each session. It does tasks that require calls to other models through OpenRouter only when it is still within its da", "context": "", "use_case": "calling other models", "switched_from": "", "story": "We're going to need default hard budget caps on pretty much everything"}, {"id": 798841872, "item_id": 49927617, "url": "https://news.ycombinator.com/item?id=49927617", "by": "adchurch", "date": "2026-10-01", "stance": "recommend", "quote": "OpenRouter does make it much easier to get started (at the cost of that variance potentially affecting quality).", "excerpt": "For our hosted version we carefully select which providers we use (we don't use OpenRouter). For the self-hosted version though, OpenRouter does make it much easier to get started (at the cost of that variance potentially affecting quality). Yes for both! We have some logic to put providers on cooldowns and/or deprioritize them.", "context": "built a model routing product", "use_case": "model provider routing for self-hosted", "switched_from": "", "story": "Show HN: Open-source model routing for coding agents at Astra-level performance"}, {"id": 798435233, "item_id": 49902202, "url": "https://news.ycombinator.com/item?id=49902202", "by": "wslh", "date": "2026-09-29", "stance": "neutral", "quote": "I didn't experience issues when I used Fable via OpenRouter", "excerpt": "> > They are calling out specific providers who release powerful models without safeguards. > Good thing Fable refuses to answer my question about how children inherit blue eyes, it was the only piece of information I needed to finish my blue-eye super-bioweapon. FWIW: I didn't experience issues when I used Fable via OpenRouter checking for security issues in code but experienced them via the Claude desktop app.", "context": "", "use_case": "accessing models", "switched_from": "", "story": "GLM-5.3 and the spread of advanced cyber capabilities"}, {"id": 798395136, "item_id": 49899696, "url": "https://news.ycombinator.com/item?id=49899696", "by": "Computer0", "date": "2026-09-29", "stance": "neutral", "quote": "I find a model/provider combination I like then pin all requests for that model to that single provider.", "excerpt": "The way I use openrouter is I find a model/provider combination I like then pin all requests for that model to that single provider.", "context": "", "use_case": "model routing with pinned provider", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}, {"id": 798186257, "item_id": 49886641, "url": "https://news.ycombinator.com/item?id=49886641", "by": "mrbonner", "date": "2026-09-29", "stance": "neutral", "quote": "the \u201cfree\u201d Ling Flash model on openrouter", "excerpt": "I have been using the \u201cfree\u201d Ling Flash model on openrouter for a bit over a month now. My work is an aside project at home building a Rust binding for an open source Zig code base library. The result is nothing impressive but also not a total failure: I have a feature parity binding to use in Rust vs. Python/java/typescript. Now, during those night and weekend sessions, I have never run into throttling issues with t", "context": "", "use_case": "accessing free models", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 798115970, "item_id": 49882248, "url": "https://news.ycombinator.com/item?id=49882248", "by": "film42", "date": "2026-09-28", "stance": "recommend", "quote": "Switched to OpenCode + OpenRouter and continued working.", "excerpt": "Working on a write-ahead log implementation, I had Opus 5.5 look to verify that it was durably writing as safely as possible. It got flagged and forced me to Opus 4.8. Switched to OpenCode + OpenRouter and continued working.", "context": "", "use_case": "model routing", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 797999986, "item_id": 49874999, "url": "https://news.ycombinator.com/item?id=49874999", "by": "babamogambo88", "date": "2026-09-28", "stance": "neutral", "quote": "sometimes other models from open router", "excerpt": "https://github.com/STRAW-HAT-DEV/monkeyDcode.git I use this coding agent with qwen 2.5b and sometimes other models from open router.. as a one piece fan lol this looks nice with all the anime stuff .. but the work it does is also promising .. I don't know if it has any issues or something .. but I feel great using it .. it almost does all the tasks I ask it to do ..", "context": "", "use_case": "accessing models", "switched_from": "", "story": "Ask HN: Which model do you use for work?"}, {"id": 797084820, "item_id": 49817801, "url": "https://news.ycombinator.com/item?id=49817801", "by": "everforward", "date": "2026-09-23", "stance": "neutral", "quote": "I do keep an OpenRouter account topped up for things that Ollama doesn\u2019t have.", "excerpt": "I use Pi and mostly open weight models. I pay for the $20/month Ollama plan and use Deepseek and GLM through that. I\u2019ve never hit the limits on it, but I tend to ask for targeted things rather than \u201cimplement a whole feature in one prompt\u201d. I do keep an OpenRouter account topped up for things that Ollama doesn\u2019t have. 99% of my usage there is embeddings, the other 1% is wanting to test some new model Ollama doesn\u2019t h", "context": "", "use_case": "embeddings and testing new models", "switched_from": "", "story": "Strands Harness"}, {"id": 797043544, "item_id": 49815221, "url": "https://news.ycombinator.com/item?id=49815221", "by": "abalashov", "date": "2026-09-23", "stance": "neutral", "quote": "with Kimi K3, via OpenRouter. This does not require sending data to China.", "excerpt": "Data points from my own usage, for whatever they're worth[1]. I'm just an individual contributor and hardly represent enterprise or large orgs: I've been on Kimi, with a little DeepSeek-V4-Pro, GLM 5.2/5.3, and MiMo thrown in, for probably about a year now. It's great here! 1) For DeepSeek, I recommend their Reasonix harness strongly, due to its alignment to DeepSeek's prefix cache. It means mostly (95%+) cache hit i", "context": "individual contributor", "use_case": "model API routing", "switched_from": "", "story": "The current balance of power in open models"}, {"id": 796953686, "item_id": 49809605, "url": "https://news.ycombinator.com/item?id=49809605", "by": "abalashov", "date": "2026-09-22", "stance": "recommend", "quote": "This does not require sending data to China.", "excerpt": "Welcome to the dark side. I've been on Kimi, with a little DeepSeek-V4-Pro, GLM 5.2/5.3, and MiMo thrown in, for probably about a year now. It's great here! For DeepSeek, I recommend their Reasonix harness strongly, due to its alignment to DeepSeek's prefix cache. It means mostly (95%+) cache hit input tokens, so very cheap large-scale code analyses and things that require mega context windows (at the cost of some at", "context": "", "use_case": "accessing Kimi K3", "switched_from": "", "story": "MiMo-v2.6-Pro: Intelligence, Performance and Price Analysis"}, {"id": 796856992, "item_id": 49803562, "url": "https://news.ycombinator.com/item?id=49803562", "by": "BeetleB", "date": "2026-09-22", "stance": "neutral", "quote": "I use OpenRouter for LLM calls", "excerpt": "Asking again (didn't get an answer in prior discussion): As there have been a lot of Jev related submissions, can someone point me to a simple guide on how I can use it? For example, say I have a script/workflow where I use OpenRouter for LLM calls, and at some point I want to do a simple classification. Can I still use OpenRouter with some Jev model...?", "context": "", "use_case": "LLM API calls", "switched_from": "", "story": "OpenAI is well positioned to fast-follow Jev"}, {"id": 796736177, "item_id": 49796011, "url": "https://news.ycombinator.com/item?id=49796011", "by": "oulu2006", "date": "2026-09-22", "stance": "warn", "quote": "I find it cheaper than paying the openrouter tax", "excerpt": "So everyone has their preferred way of doing it. 1. I go direct to source, i.e. DS platform, I find it cheaper than paying the openrouter tax -- I also switch it up a bit 2. I built a local LLM router, that I update with new profiles that have my preferred provider of the week (lowest token costs/speed) with fallbacks, like mimo --> DS4 etc.. if there is overloading, 3. I use 3 diff harnesses, CC/Codex + Opencode -- ", "context": "", "use_case": "model routing", "switched_from": "", "story": "Elevated errors for multiple models \u2013 Resolved"}, {"id": 796699777, "item_id": 49793736, "url": "https://news.ycombinator.com/item?id=49793736", "by": "lanyard-textile", "date": "2026-09-21", "stance": "neutral", "quote": "Fortunately it's on OpenRouter for everyone now", "excerpt": "Thank you ;) +1 on the animation. Thanks for the feedback. Fortunately it's on OpenRouter for everyone now :) I think. I use my Jev BYOK, but I assume you don't need the key.", "context": "", "use_case": "model access", "switched_from": "", "story": "Show HN: Will Jev pull the lever in the trolley problem?"}, {"id": 796680672, "item_id": 49792542, "url": "https://news.ycombinator.com/item?id=49792542", "by": "RussianCow", "date": "2026-09-21", "stance": "recommend", "quote": "I regularly get ~90-95% cache hit rates.", "excerpt": "For caching, only if you don't specify your preferred providers and let OpenRouter route each request itself. I have stuff like this in my OpenCode config for each model I use and I regularly get ~90-95% cache hit rates. \"order\": [\"relace\", \"coreweave\", \"novita\", \"baseten\", \"together\"], \"allow_fallbacks\": false It still won't be quite as high as you'd get by just using DeepSeek because occasionally a request will fai", "context": "", "use_case": "model routing with provider order", "switched_from": "", "story": "Grok 4.7"}, {"id": 795154673, "item_id": 49697167, "url": "https://news.ycombinator.com/item?id=49697167", "by": "Roark66", "date": "2026-09-14", "stance": "warn", "quote": "Openrouter is not ideal, because you don't know who they send your traffic to", "excerpt": "Openrouter is not ideal, because you don't know who they send your traffic to and there are rumours of vendors cheating by providing quantized models. I'm very happy with Deepinfra. Less model coverage, but good prices and quite fast. However I have to caution you about one thing. No one will give you as many input tokens for so little money as Claude Max x5 (maybe x20 too, I use x5). I tend to use 1.1B to 1.4B a wee", "context": "", "use_case": "", "switched_from": "", "story": "Big AI sets out its terms for regulatory capture"}, {"id": 794625216, "item_id": 49664076, "url": "https://news.ycombinator.com/item?id=49664076", "by": "pessimizer", "date": "2026-09-11", "stance": "warn", "quote": "If it doesn't do that, then what is it good for? If I have to permanently pin the one provider who doesn't ignore what I've asked, why shouldn't I just deal with them directly?", "excerpt": "> Parent is complaining about the auto-provider chosing, not that all providers are unreliable. Parent is claiming that the choosing is unreliable. If I rely on a provider to bring me tuna to some spec, but they get it from many different fishermen, it doesn't mean that the tuna doesn't have meet the spec. The complaint is that they're given a bunch of knobs that simply don't work with providers that they could be sw", "context": "", "use_case": "LLM provider routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 794512976, "item_id": 49657061, "url": "https://news.ycombinator.com/item?id=49657061", "by": "embedding-shape", "date": "2026-09-11", "stance": "recommend", "quote": "This is quite literally the point of OpenRouter. A unified interface, so you can easily switch providers without changing a ton of code", "excerpt": "> just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. But that's the intention right? Even the name implies they just send stuff around for you, and if you want to control the routing, you'd lock down providers. I don't see how they could build what they wanted to build, and not have it end up unreliable if you freely round-robin between providers, i", "context": "", "use_case": "unified LLM provider interface", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 794501328, "item_id": 49656333, "url": "https://news.ycombinator.com/item?id=49656333", "by": "joshstrange", "date": "2026-09-11", "stance": "warn", "quote": "It\u2019s just incredibly unreliable and you are forced to pin providers", "excerpt": "This squares with my, much much, smaller OpenRouter usage. It\u2019s just incredibly unreliable and you are forced to pin providers and even then it can be a crapshoot as the author found. OpenRouter sells the idea of swapping being commodity providers but it couldn\u2019t be further from the truth. Provider A is often not swappable for B or C (again, as this author found). It can be crazy-making as you sit there thinking \u201cOpe", "context": "", "use_case": "LLM routing", "switched_from": "", "story": "So you want to use OpenRouter?"}, {"id": 790656816, "item_id": 49416051, "url": "https://news.ycombinator.com/item?id=49416051", "by": "sznio", "date": "2026-08-24", "stance": "recommend", "quote": "I just put $30 on openrouter, switched to Pi, and I finally have a calm mind.", "excerpt": "In my case it's been kind of unhealthy, just constantly waiting for the token limit to reset, always feeling like I'm wasting a resource if I'm not using subscription right now. I just put $30 on openrouter, switched to Pi, and I finally have a calm mind. Since I actually pay per request I want to maximize efficiency rather than utilization", "context": "", "use_case": "pay-per-request model access", "switched_from": "Claude", "story": "Anthropic's best AI model struggles to attract users as cheaper tools thrive"}, {"id": 790570098, "item_id": 49410631, "url": "https://news.ycombinator.com/item?id=49410631", "by": "microtonal", "date": "2026-08-23", "stance": "recommend", "quote": "Use OpenRouter or HF Inference Providers in between and you can effortlessly switch between models and providers.", "excerpt": "peace of mind versus using some Chinese cloud These are open weight models (GLM-5.3 soon too). You can run them on the Together AIs or Firework AIs of this world. Use OpenRouter or HF Inference Providers in between and you can effortlessly switch between models and providers. I have been using GLM and Kimi models the last few months mixed with the latest Anthropic models and for my daily work there is barely a differ", "context": "", "use_case": "switching models and providers", "switched_from": "", "story": "GLM-5.3 (open-weight) beat Anthropic/OpenAI models \u2013 for 1/5 the cost"}, {"id": 789676544, "item_id": 49354784, "url": "https://news.ycombinator.com/item?id=49354784", "by": "oceanplexian", "date": "2026-08-18", "stance": "recommend", "quote": "$200 buys a lot of tokens on OpenRouter.", "excerpt": "I use the $200 plan w/ Anthropic and run out of tokens half way through the week and supposedly they are progressively reducing the limits on all their subs even further. At some point I will switch, $200 buys a lot of tokens on OpenRouter.", "context": "pays for the $200 Anthropic plan", "use_case": "cheaper LLM tokens", "switched_from": "", "story": "GLM-5.3 Artificial Analysis Benchmarks"}, {"id": 789449120, "item_id": 49340570, "url": "https://news.ycombinator.com/item?id=49340570", "by": "PreownedPlaid", "date": "2026-08-17", "stance": "recommend", "quote": "Switching models without opening a new vendor account is the thing I use it for.", "excerpt": "The product people actually pay for is one bill and one API key. Routing is the marketing. Switching models without opening a new vendor account is the thing I use it for.", "context": "", "use_case": "switching models via one API key", "switched_from": "", "story": "Stripe Clinches over $7B Deal to Buy AI Firm OpenRouter"}, {"id": 786449057, "item_id": 49153066, "url": "https://news.ycombinator.com/item?id=49153066", "by": "thecopy", "date": "2026-08-03", "stance": "warn", "quote": "Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern.", "excerpt": "I agree with parent. OpenRouter might be cheaper list-price, but i have been using 10$ on DS platform since April/May, still have 2$ left. Using OpenRouter i depleted the same dollar-amount in a 1-2 weeks with same usage pattern. No idea why.", "context": "", "use_case": "LLM API access", "switched_from": "", "story": "Qwen3.8-Max: A New Bar for Coding and Cowork"}, {"id": 781022736, "item_id": 48813921, "url": "https://news.ycombinator.com/item?id=48813921", "by": "KetoManx64", "date": "2026-07-06", "stance": "recommend", "quote": "I use OpenRoutet which lets you switch between providers (Anthropic, ChatGPT, Z-AI) whenever you want.", "excerpt": "Most of the AI companies have OpenAi compatible API's, so you just get a subscription from another provider and change the URL that your LLM Agent Harness uses to talk to the AI. I use OpenRoutet which lets you switch between providers (Anthropic, ChatGPT, Z-AI) whenever you want. Sometimes I'll have two different models from different providers evaluate each other's answers.", "context": "", "use_case": "switching LLM providers", "switched_from": "", "story": "GLM 5.2 and the coming AI margin collapse"}, {"id": 779068960, "item_id": 48691810, "url": "https://news.ycombinator.com/item?id=48691810", "by": "stanac", "date": "2026-06-26", "stance": "recommend", "quote": "I switched to openrouter and have enough prepaid credits to last me for months with Chinese models.", "excerpt": "Last time I used subscription I hit a usage limit in 20 minutes (gemini). I switched to openrouter and have enough prepaid credits to last me for months with Chinese models. I spent about $30 in last two months.", "context": "", "use_case": "pay-per-use model access", "switched_from": "Gemini", "story": "U.S. government will decide who gets to use GPT-5.6"}, {"id": 778711984, "item_id": 48669499, "url": "https://news.ycombinator.com/item?id=48669499", "by": "tacomagick", "date": "2026-06-24", "stance": "warn", "quote": "Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API", "excerpt": "For my case Openrouter breaks Deepseek caching and charges me multiple times over what I pay for Deepseek's API, with 2$ I was able to get around 120M tokens from deepseek easily when Openrouter could only barely do 250k", "context": "", "use_case": "accessing DeepSeek models", "switched_from": "", "story": "GLM-5.2 is the step change for open agents"}]}