{"slug": "runpod", "name": "RunPod", "category": "infra", "category_name": "Infrastructure and DevOps", "description": "", "recommenders": 5, "commenters": 15, "warners": 1, "score": 5.928, "rising": -0.31, "first": "2025-12-03", "latest": "2026-08-26", "why": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte.", "why_is_quote": true, "has_verdict": false, "sentiment": {"recommend": 5, "neutral": 8, "compare": 1, "warn": 1}, "spark": [0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 0], "listed": true, "loved": [], "caveat": "", "short_quote": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte.", "kind": "", "icon": true, "domain": "runpod.io", "github_repo": null, "official_url": "https://www.runpod.io", "pricing": "", "aliases": ["RunPod", "Runpod"], "verdict": null, "praise": [{"label": "", "summary": "", "quotes": [{"id": 791310545, "item_id": 49456909, "url": "https://news.ycombinator.com/item?id=49456909", "by": "girvo", "date": "2026-08-26", "version": null, "stance": "recommend", "quote": "you're better off using Vast or Runpod or whatever to rent cloud compute. Much faster and cheap as hell", "excerpt": "I love my Spark-like, but even for training you're better off using Vast or Runpod or whatever to rent cloud compute. Much faster and cheap as hell, to be honest. I do set up my initial runs and likes like quantisation-aware-distillation on my Spark-like to test it out and get it working, so it has value! But its not \"worth\" it other than its fun hardware to tinker with, IMO.", "context": "has a Spark-like local machine", "use_case": "renting GPUs for training", "switched_from": "", "story": "GLM-5.3-Flash"}, {"id": 785964882, "item_id": 49122805, "url": "https://news.ycombinator.com/item?id=49122805", "by": "JamieAlethieum", "date": "2026-07-31", "version": null, "stance": "recommend", "quote": "There is also Runpod and a few more.", "excerpt": "There are a couple of options. One good way to find inference providers for open models is through hugging face (https://huggingface.co). you can select a model and see which inference providers serve it. you can even access it through hugging face. If you just wanted to test a model or have super light work you can get some free access to alot of open source models through nvidia (https://build.nvidia.com). There is", "context": "", "use_case": "LLM inference", "switched_from": "", "story": "Ask HN: What are you using for LLM inference in production?"}, {"id": 774588720, "item_id": 48411795, "url": "https://news.ycombinator.com/item?id=48411795", "by": "theletterf", "date": "2026-06-05", "version": null, "stance": "recommend", "quote": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte.", "excerpt": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte. Stanford's NLP course recommends them: https://cs336.stanford.edu/", "context": "", "use_case": "Renting GPUs for fine-tuning", "switched_from": "", "story": "Fine-tuning an LLM to write docs like it's 1995"}, {"id": 755715618, "item_id": 47232226, "url": "https://news.ycombinator.com/item?id=47232226", "by": "enginous", "date": "2026-03-03", "version": null, "stance": "recommend", "quote": "I'd rather use services like RunPod for training jobs", "excerpt": "One caveat to alerts (and automatically acting on alerts) is that there are delays[0] between costs being incurred and alerted. I can't find a Google source for what the delay is, but a source online say it could be \"24 hours [to] a few days.\"[1] This has been a major reason why I reach for OpenAI models before Gemini, but also why I'd rather use services like RunPod for training jobs. For a small boostrapped company", "context": "small bootstrapped company", "use_case": "GPU training jobs", "switched_from": "", "story": "Stolen Gemini API key racks up $82,000 in 48 hours"}, {"id": 747173510, "item_id": 46698344, "url": "https://news.ycombinator.com/item?id=46698344", "by": "dvfjsdhgfv", "date": "2026-01-20", "version": null, "stance": "recommend", "quote": "I'm very happy with the results", "excerpt": "I pay $20 for ChatGPT, I ask it to criticize my code and ideas. Sometimes it's useful, sometimes it says bullshit. For a few months I used Gemini Pro, there was a period when it was better than OpenAI's model but they did something and now it's worse even though it answers faster so I cancelled my Google One subscription. I tried Claude Code over a few weekends, it definitely can do tiny projects quickly but I work i", "context": "", "use_case": "Renting GPUs for ML", "switched_from": "", "story": "Majority of CEOs report zero payoff from AI splurge"}]}], "caveats": [{"label": "", "summary": "", "quotes": [{"id": 757686581, "item_id": 47355411, "url": "https://news.ycombinator.com/item?id=47355411", "by": "vshah1016", "date": "2026-03-12", "version": null, "stance": "warn", "quote": "cheap-but-DIY (Modal, RunPod \u2014 you configure vLLM yourself and deal with slow cold starts)", "excerpt": "Hey HN \u2014 I\u2019m Veer and my cofounder is Suryaa. We're building Cumulus Labs (YC W26), and we're releasing our latest product IonRouter (https://ionrouter.io/), an inference API for open-source and fine tuned models. You swap in our base URL, keep your existing OpenAI client code, and get access to any model (open source or finetuned to you) running on our own inference engine. The problem we kept running into: every in", "context": "", "use_case": "", "switched_from": "", "story": "IonRouter (YC W26) \u2013 High throughput, low cost inference"}]}], "who": [{"by": "girvo", "context": "has a Spark-like local machine", "url": "https://news.ycombinator.com/item?id=49456909", "date": "2026-08-26"}, {"by": "Danau5tin", "context": "RL practitioner", "url": "https://news.ycombinator.com/item?id=48907077", "date": "2026-07-14"}, {"by": "enginous", "context": "small bootstrapped company", "url": "https://news.ycombinator.com/item?id=47232226", "date": "2026-03-03"}, {"by": "scylx", "context": "runs print studio Axisophy", "url": "https://news.ycombinator.com/item?id=47017106", "date": "2026-02-14"}], "tags": [], "use_cases": ["gpu cloud hosting", "cloud gpu rental", "renting gpus for training", "llm inference", "gpu compute", "renting gpus for fine-tuning"], "compared": [{"slug": "vast-ai", "name": "Vast.ai", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "gemini", "name": "Gemini", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=49389952", "title": "How We Made a Text-to-Speech Model Respond in Sub-50 ms", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=49449507", "title": "GLM-5.3-Flash", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49121047", "title": "Ask HN: What are you using for LLM inference in production?", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48905919", "title": "Show HN: I RL-trained an agent that trains models with RL (for \u2013$1.3k)", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48408442", "title": "Fine-tuning an LLM to write docs like it's 1995", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47792764", "title": "Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All", "mentions": 1}], "sentiment_by_month": {"2025-11": {}, "2025-12": {"compare": 1}, "2026-01": {"recommend": 1, "neutral": 1}, "2026-02": {"neutral": 1}, "2026-03": {"recommend": 1, "neutral": 2, "warn": 1}, "2026-04": {"neutral": 2}, "2026-05": {}, "2026-06": {"recommend": 1}, "2026-07": {"neutral": 1, "recommend": 1}, "2026-08": {"neutral": 1, "recommend": 1}, "2026-09": {}, "2026-10": {}}, "versions": [], "version_unnamed": 0, "family_ref": null, "runs_on": null, "lines": [], "quotes": [{"id": 791310545, "item_id": 49456909, "url": "https://news.ycombinator.com/item?id=49456909", "by": "girvo", "date": "2026-08-26", "version": null, "stance": "recommend", "quote": "you're better off using Vast or Runpod or whatever to rent cloud compute. Much faster and cheap as hell", "excerpt": "I love my Spark-like, but even for training you're better off using Vast or Runpod or whatever to rent cloud compute. Much faster and cheap as hell, to be honest. I do set up my initial runs and likes like quantisation-aware-distillation on my Spark-like to test it out and get it working, so it has value! But its not \"worth\" it other than its fun hardware to tinker with, IMO.", "context": "has a Spark-like local machine", "use_case": "renting GPUs for training", "switched_from": "", "story": "GLM-5.3-Flash"}, {"id": 790375104, "item_id": 49398444, "url": "https://news.ycombinator.com/item?id=49398444", "by": "totetsu", "date": "2026-08-22", "version": null, "stance": "neutral", "quote": "I tried it out on runpod for about $4", "excerpt": "I tried it out on runpod for about $4 using this template. you might need to override the environment varrs and put in a hugging face token for it to pull model. https://console.runpod.io/hub/template/pahlszv0ul?ref=0xiswpjk", "context": "", "use_case": "GPU cloud hosting", "switched_from": "", "story": "How We Made a Text-to-Speech Model Respond in Sub-50 ms"}, {"id": 790373104, "item_id": 49398319, "url": "https://news.ycombinator.com/item?id=49398319", "by": "totetsu", "date": "2026-08-22", "version": null, "stance": "neutral", "quote": "I just tried it out on runpod", "excerpt": "I just tried it out on runpod [1] on the lady of shalott. Its got a funny sing song affect https://pastewaves.com/player/cdc4693a-a818-4d34-8566-f0207d502127 that one took maybe 12.677831s to generate.. DNS: 0.003747s Connect: 0.044571s TLS: 0.115428s First byte: 12.677831s Total curl: 26.268993s HTTP status: 200 Downloaded: 7695404 bytes [1] with this template https://console.runpod.io/hub/template/pahlszv0ul?ref=0x", "context": "", "use_case": "GPU cloud hosting", "switched_from": "", "story": "How We Made a Text-to-Speech Model Respond in Sub-50 ms"}, {"id": 785964882, "item_id": 49122805, "url": "https://news.ycombinator.com/item?id=49122805", "by": "JamieAlethieum", "date": "2026-07-31", "version": null, "stance": "recommend", "quote": "There is also Runpod and a few more.", "excerpt": "There are a couple of options. One good way to find inference providers for open models is through hugging face (https://huggingface.co). you can select a model and see which inference providers serve it. you can even access it through hugging face. If you just wanted to test a model or have super light work you can get some free access to alot of open source models through nvidia (https://build.nvidia.com). There is", "context": "", "use_case": "LLM inference", "switched_from": "", "story": "Ask HN: What are you using for LLM inference in production?"}, {"id": 782513235, "item_id": 48907077, "url": "https://news.ycombinator.com/item?id=48907077", "by": "Danau5tin", "date": "2026-07-14", "version": null, "stance": "neutral", "quote": "the stack of Tinker, Prime-RL & Runpod - all of which I know well", "excerpt": "I chose the key technical decision and direction (such as the system architecture, the tasks to train on, the stack of Tinker, Prime-RL & Runpod - all of which I know well) etc. The problems it would do well on are training small agentic (multi-turn, tool use) task based models using the prime-rl stack, which are close to the distribution trained upon. It would likely not transfer to other training frameworks such as", "context": "RL practitioner", "use_case": "GPU compute", "switched_from": "", "story": "Show HN: I RL-trained an agent that trains models with RL (for \u2013$1.3k)"}, {"id": 774588720, "item_id": 48411795, "url": "https://news.ycombinator.com/item?id=48411795", "by": "theletterf", "date": "2026-06-05", "version": null, "stance": "recommend", "quote": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte.", "excerpt": "Try Runpod or similar services: you can fine-tune stuff for the price of a latte. Stanford's NLP course recommends them: https://cs336.stanford.edu/", "context": "", "use_case": "Renting GPUs for fine-tuning", "switched_from": "", "story": "Fine-tuning an LLM to write docs like it's 1995"}, {"id": 764954752, "item_id": 47809672, "url": "https://news.ycombinator.com/item?id=47809672", "by": "Tamitami2000", "date": "2026-04-17", "version": null, "stance": "neutral", "quote": "I used runpod before, but they don't have systems like that.", "excerpt": "Can you tell me where? I used runpod before, but they don't have systems like that.", "context": "", "use_case": "Cloud GPU rental", "switched_from": "", "story": "Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All"}, {"id": 764123330, "item_id": 47757708, "url": "https://news.ycombinator.com/item?id=47757708", "by": "christianh", "date": "2026-04-13", "version": null, "stance": "neutral", "quote": "The heavy lifting happens in Runpod.", "excerpt": "I\u2019m working on https://poser.pro, a ski technique feedback tool for skiers that want to improve their skills. How? The user (skier!) uploads a short ski clip to Poser, lets Poser analyze, and then gets their results. For now, you get video outputs: head-tracking and skeleton overlay. I\u2019ve built a a lot of extra stuff (animated 3D body, turn-detection, balance-, steering-, pressure-, and edging-metrics, etc.), I\u2019m jus", "context": "", "use_case": "", "switched_from": "", "story": "Ask HN: What Are You Working On? (April 2026)"}, {"id": 761556290, "item_id": 47597268, "url": "https://news.ycombinator.com/item?id=47597268", "by": "freakynit", "date": "2026-03-31", "version": null, "stance": "neutral", "quote": "https://ofo1j9j6qh20a8-80.proxy.runpod.net", "excerpt": "Open access for next 5 hours (8GiB model, running on RTX 3090) or until server crashes or the this spot instance gets taken away :) => https://ofo1j9j6qh20a8-80.proxy.runpod.net ./build/bin/llama-server \\ -m ../Bonsai-8B.gguf \\ -ngl 999 \\ --flash-attn on \\ --host 0.0.0.0 \\ --port 80 \\ --ctx-size 65500 \\ --batch-size 512 \\ --ubatch-size 512 \\ --parallel 5 \\ --cont-batching \\ --threads 8 \\ --threads-batch 8 \\ --cache-t", "context": "", "use_case": "GPU hosting for demo", "switched_from": "", "story": "Show HN: 1-Bit Bonsai, the First Commercially Viable 1-Bit LLMs"}, {"id": 757686581, "item_id": 47355411, "url": "https://news.ycombinator.com/item?id=47355411", "by": "vshah1016", "date": "2026-03-12", "version": null, "stance": "warn", "quote": "cheap-but-DIY (Modal, RunPod \u2014 you configure vLLM yourself and deal with slow cold starts)", "excerpt": "Hey HN \u2014 I\u2019m Veer and my cofounder is Suryaa. We're building Cumulus Labs (YC W26), and we're releasing our latest product IonRouter (https://ionrouter.io/), an inference API for open-source and fine tuned models. You swap in our base URL, keep your existing OpenAI client code, and get access to any model (open source or finetuned to you) running on our own inference engine. The problem we kept running into: every in", "context": "", "use_case": "", "switched_from": "", "story": "IonRouter (YC W26) \u2013 High throughput, low cost inference"}, {"id": 756235697, "item_id": 47264731, "url": "https://news.ycombinator.com/item?id=47264731", "by": "k2so", "date": "2026-03-05", "version": null, "stance": "neutral", "quote": "had to pick between three H100 variants on Runpod", "excerpt": "Author here. I wanted to train Nanochat d26 to GPT-2 level and had to pick between three H100 variants on Runpod. SXM was the most expensive per hour but cheapest to finish: SXM: 702ms/step - ~$37 (using vast.ai) PCIe: 1,412ms/step - ~$112 (runpod) NVL: 2,032ms/step - ~$181 (runpod) My first SXM run hit 1,295ms. Barely faster than PCIe. Nsight OS runtime summary led me to suspect CPU starvation. I found a higher vCPU", "context": "", "use_case": "", "switched_from": "", "story": "Which H100 Instance to Train Nanochat \u2013 Benchmarking PCIe, SXM, and NVL"}, {"id": 755715618, "item_id": 47232226, "url": "https://news.ycombinator.com/item?id=47232226", "by": "enginous", "date": "2026-03-03", "version": null, "stance": "recommend", "quote": "I'd rather use services like RunPod for training jobs", "excerpt": "One caveat to alerts (and automatically acting on alerts) is that there are delays[0] between costs being incurred and alerted. I can't find a Google source for what the delay is, but a source online say it could be \"24 hours [to] a few days.\"[1] This has been a major reason why I reach for OpenAI models before Gemini, but also why I'd rather use services like RunPod for training jobs. For a small boostrapped company", "context": "small bootstrapped company", "use_case": "GPU training jobs", "switched_from": "", "story": "Stolen Gemini API key racks up $82,000 in 48 hours"}, {"id": 752273696, "item_id": 47017106, "url": "https://news.ycombinator.com/item?id=47017106", "by": "scylx", "date": "2026-02-14", "version": null, "stance": "neutral", "quote": "The renders are done on A100s and RTX 4090s via RunPod.", "excerpt": "I posted my print studio Axisophy on here years ago under a different name. I've recently relaunched it and have been spending most of my GPU time on the Mersenne series - strange attractors, phase portraits, Weyl symmetry projections from E8 root systems, that kind of thing. The renders are done on A100s and RTX 4090s via RunPod. Each piece is computed at 200+ megapixels (enough for 40-inch square prints at 300dpi) ", "context": "runs print studio Axisophy", "use_case": "GPU rendering compute", "switched_from": "", "story": "Rendering attractors at 200 megapixels on A100s"}, {"id": 747828065, "item_id": 46739254, "url": "https://news.ycombinator.com/item?id=46739254", "by": "ecmulli", "date": "2026-01-23", "version": null, "stance": "neutral", "quote": "via Runpod - very cool!", "excerpt": "I dont have a big enough GPU but I was able to play around with the model using this plugin https://github.com/daydreamlive/scope-overworld via Runpod - very cool!", "context": "", "use_case": "cloud GPU rental", "switched_from": "", "story": "Waypoint-1: Real-Time Interactive Video Diffusion from Overworld"}, {"id": 747173510, "item_id": 46698344, "url": "https://news.ycombinator.com/item?id=46698344", "by": "dvfjsdhgfv", "date": "2026-01-20", "version": null, "stance": "recommend", "quote": "I'm very happy with the results", "excerpt": "I pay $20 for ChatGPT, I ask it to criticize my code and ideas. Sometimes it's useful, sometimes it says bullshit. For a few months I used Gemini Pro, there was a period when it was better than OpenAI's model but they did something and now it's worse even though it answers faster so I cancelled my Google One subscription. I tried Claude Code over a few weekends, it definitely can do tiny projects quickly but I work i", "context": "", "use_case": "Renting GPUs for ML", "switched_from": "", "story": "Majority of CEOs report zero payoff from AI splurge"}, {"id": 738246160, "item_id": 46140385, "url": "https://news.ycombinator.com/item?id=46140385", "by": "michaelt", "date": "2025-12-03", "version": null, "stance": "compare", "quote": "And Runpod is one of the cheaper cloud providers", "excerpt": "On \"runpod community cloud\" renting a 5090 costs $0.69/hour [1] and it consumes about $0.10/hour electricity, if running at full power and paying $0.20/kWh. On Amazon, buying a 5090 costs $3000 [2] That's a payback time of 212 days. And Runpod is one of the cheaper cloud providers; for the GPUs I compared, EC2 was twice the price for an on-demand instance. Rental prices for GPUs are pretty darn high. [1] https://www.", "context": "", "use_case": "Renting GPU compute", "switched_from": "", "story": "Are we repeating the telecoms crash with AI datacenters?"}]}