{"slug": "deepseek-v4-flash", "name": "DeepSeek V4 Flash", "category": "ai", "category_name": "AI models and assistants", "description": "DeepSeek's fast, low-cost open-weight large language model, offered via API and for local use.", "recommenders": 153, "commenters": 200, "warners": 5, "score": 148.813, "rising": 94.96, "first": "2026-04-24", "latest": "2026-10-29", "why": "HN commenters like it as a fast, very cheap daily driver for coding, often using it to implement plans written by pricier frontier models.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 171, "neutral": 38, "compare": 12, "warn": 5}, "spark": [0, 0, 0, 0, 0, 1, 6, 28, 19, 45, 33, 39], "listed": true, "loved": ["Low cost", "Speed"], "caveat": "Below frontier", "short_quote": "I love DeepSeek V4 Flash and I use it extensively, it's so cheap I can use it all day", "icon": true, "domain": "artificialanalysis.ai", "github_repo": null, "official_url": "https://www.deepseek.com", "pricing": "open source", "aliases": ["DeepSeek V4 Flash", "DeepSeek Flash", "DeepSeek V4.1 Flash", "DeepSeek 4.1 Flash", "DeepSeek Flash 4.1", "DeepSeek v4.1 Flash", "DeepSeek-V4-Flash", "DeepSeek 4 Flash", "DeepSeek V4 Flash 0731", "deepseek-v4-flash-0731", "DeepSeek v4 Flash", "DeepSeek flash", "DeepSeek-V4.1-Flash"], "verdict": [{"text": "HN commenters like it as a fast, very cheap daily driver for coding", "cites": [{"id": 804330976, "item_id": 50270686, "url": "https://news.ycombinator.com/item?id=50270686", "by": "piterrro", "date": "2026-10-28", "stance": "recommend", "quote": "I'm still using DS-flash - it beats frontier model on speed and cost keeping good enough code quality.", "excerpt": "I'm still using DS-flash - it beats frontier model on speed and cost keeping good enough code quality. It depends on the harness and constraints you give to the model but for most of the tasks its a great fit, I have not switched from DS since couple of weeks already (previously I would switch to Opus for some heavier work but I stopped recently). My harness consists of multi-level AGENTS.md and SOUL.md containing \"g", "context": "", "use_case": "coding tasks", "switched_from": "Opus", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 803033600, "item_id": 50189600, "url": "https://news.ycombinator.com/item?id=50189600", "by": "brazukadev", "date": "2026-10-22", "stance": "recommend", "quote": "It is terribly fast! And smart. And cheap.", "excerpt": "I used to think that until I started using DeepSeek V4.1 Flash 1017 for the past 2 days (what a bad name). It is terribly fast! And smart. And cheap.", "context": "", "use_case": "", "switched_from": "", "story": "Token-maxxing is dead, OpenRouter's token usage plateaus"}, {"id": 800204064, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "orchestrator, tester, subagent", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800315056, "item_id": 50019691, "url": "https://news.ycombinator.com/item?id=50019691", "by": "Primer81", "date": "2026-10-09", "stance": "recommend", "quote": "Its a daily driver for sure. not even worth considering anything else", "excerpt": "I've been waiting for something like this article for a month. The other AI companies are being left in the dust. I rip about 200 million tokens at least a day and spend maybe $3 with deepseek flash v4.1 for self hosted related coding tasks for ~20 projects i work on / maintain in parallel. Its a daily driver for sure. not even worth considering anything else, but maybe a locally run model on my GPU at the moment.", "context": "maintains ~20 projects", "use_case": "self-hosted related coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 803966624, "item_id": 50247914, "url": "https://news.ycombinator.com/item?id=50247914", "by": "joshheitzman", "date": "2026-10-26", "stance": "recommend", "quote": "I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "excerpt": "As a counter-example, I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "context": "", "use_case": "daily coding on multiple projects", "switched_from": "", "story": "Anthropic S-1"}]}, {"text": ", often using it to implement plans written by pricier frontier models.", "cites": [{"id": 800250993, "item_id": 50015687, "url": "https://news.ycombinator.com/item?id=50015687", "by": "yuhmahp", "date": "2026-10-09", "stance": "recommend", "quote": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "excerpt": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "context": "", "use_case": "general coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 795660832, "item_id": 49728802, "url": "https://news.ycombinator.com/item?id=49728802", "by": "surgical_fire", "date": "2026-09-16", "stance": "recommend", "quote": "I love DS flash, it is an amazing workhorse to implement plans created by more robust models", "excerpt": "What I find surprising is that DS Flash can do it at all. I love DS flash, it is an amazing workhorse to implement plans created by more robust models (such as GLM). But a more fair comparison would be of DS Flash with GLM Flash.", "context": "", "use_case": "implementing plans", "switched_from": "", "story": "DeepSeek v4.1 Flash Is Now Our Best Hacking Model"}, {"id": 800197696, "item_id": 50012356, "url": "https://news.ycombinator.com/item?id=50012356", "by": "apitman", "date": "2026-10-08", "stance": "recommend", "quote": "despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks", "excerpt": "> With my OpenCode Go sub of $10/month, DeepSeek is basically unlimited My OpenCode Go monthly window was scheduled to reset this morning. It was sitting at 22% used despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks (I use gpt-6.1-sol high for planning/orchestration). I had 1.5 hours left so I fired up first 10, then 20, and finally 50 concurrent subagents all working on re", "context": "", "use_case": "implementation agent, subagents", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 802077776, "item_id": 50129861, "url": "https://news.ycombinator.com/item?id=50129861", "by": "gregwebs", "date": "2026-10-17", "stance": "recommend", "quote": "I am going back to DeepSeek for writing code.", "excerpt": "The company's own news page: https://api-docs.deepseek.com/news/news261017 They are showing this new version scoring significantly higher across multiple coding benchmarks, beating Sol 5.6 (6.1 is the latest) across 10/11 benchmarks. I am going back to DeepSeek for writing code. I was trying out using Sol and liking the result, but this should close a lot of that gap. I use Opus/Sol for an implementation plan and cod", "context": "", "use_case": "writing code", "switched_from": "Sol", "story": "DeepSeek-v4.1-Flash-1017"}]}, {"text": " The most common caveat is that it trails frontier models on harder design, explanation and debugging work", "cites": [{"id": 800204064, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "orchestrator, tester, subagent", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800403126, "item_id": 50025195, "url": "https://news.ycombinator.com/item?id=50025195", "by": "aftbit", "date": "2026-10-09", "stance": "warn", "quote": "they're more likely to go down rabbit holes, confidently assert falsehoods, or fix bugs with changes to my test harness rather than my code.", "excerpt": "I have had good luck rotating between the three tiers of GPT 5.6 with occasional jumps up to Astra or Fable. Most of my work is with GPT-5.6-Sol. Simple tasks like data extraction or trivial refactors (rename this variable etc) I often push down to Terra or Luna, or even to self-hosted Gemma4:31B. Very tricky stuff, like planning a new feature, design review, or code review of a complex change across multiple reposit", "context": "", "use_case": "easy coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800234272, "item_id": 50014642, "url": "https://news.ycombinator.com/item?id=50014642", "by": "ltbarcly3", "date": "2026-10-09", "stance": "warn", "quote": "DeepSeek 4.1 Flash kindof sucks.", "excerpt": "DeepSeek 4.1 Flash kindof sucks. I used it a bunch and it kindof sucks. I don't know if they are gaming benchmarks or what. Luna is on par in benchmarks and my personal experience is Luna is better for what I do, and Luna is cheaper. Comparing Deepseek 4.1 flash to Opus is just ludicrous. https://artificialanalysis.ai/models/releases/comparisons?compare=gpt-6-luna%2Cdeepseek-v4-1-flash", "context": "", "use_case": "", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 798302640, "item_id": 49893915, "url": "https://news.ycombinator.com/item?id=49893915", "by": "dools", "date": "2026-09-29", "stance": "recommend", "quote": "Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do.", "excerpt": "I go for weeks without topping up my Deepseek balance, and have to top up my OpenAI balance at least every week. GPT 6 Luna closed the gap significantly for sure (it seems to be about twice as expensive as DS v4.1 Flash), but Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do. Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt, ", "context": "", "use_case": "planning and implementation", "switched_from": "", "story": "Sonnet 5.5"}]}, {"text": ", and its chattiness can eat into the per-token savings.", "cites": [{"id": 800198496, "item_id": 50012406, "url": "https://news.ycombinator.com/item?id=50012406", "by": "rapind", "date": "2026-10-08", "stance": "recommend", "quote": "I do think it's the most viable alternative I've seen so far", "excerpt": "It's really not cheaper than frontier subscriptions. It's getting closer, and it's a great model, but it is not more value per task than the frontier subscriptions. Don't be swayed by the token costs, it's very chatty, like 3x more tokens for the same task as sol. I used dsf 4.1 full time for about a week. It blows frontier API pricing out of the water, but again, look at cost per task, not token usage. Still easily ", "context": "", "use_case": "full-time coding work", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 798302640, "item_id": 49893915, "url": "https://news.ycombinator.com/item?id=49893915", "by": "dools", "date": "2026-09-29", "stance": "recommend", "quote": "Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do.", "excerpt": "I go for weeks without topping up my Deepseek balance, and have to top up my OpenAI balance at least every week. GPT 6 Luna closed the gap significantly for sure (it seems to be about twice as expensive as DS v4.1 Flash), but Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do. Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt, ", "context": "", "use_case": "planning and implementation", "switched_from": "", "story": "Sonnet 5.5"}]}], "praise": [{"label": "Low cost", "summary": "Heavy all-day use costs only a few dollars, far below frontier subscriptions or APIs.", "quotes": [{"id": 800347056, "item_id": 50021691, "url": "https://news.ycombinator.com/item?id=50021691", "by": "epolanski", "date": "2026-10-09", "stance": "recommend", "quote": "I use DS 4.1 flash it all the time, i struggle to spend more than 1 euro per day on it", "excerpt": "> It's really not cheaper than frontier subscriptions. I use DS 4.1 flash it all the time, i struggle to spend more than 1 euro per day on it, even working all day long.", "context": "", "use_case": "daily coding work", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800315056, "item_id": 50019691, "url": "https://news.ycombinator.com/item?id=50019691", "by": "Primer81", "date": "2026-10-09", "stance": "recommend", "quote": "Its a daily driver for sure. not even worth considering anything else", "excerpt": "I've been waiting for something like this article for a month. The other AI companies are being left in the dust. I rip about 200 million tokens at least a day and spend maybe $3 with deepseek flash v4.1 for self hosted related coding tasks for ~20 projects i work on / maintain in parallel. Its a daily driver for sure. not even worth considering anything else, but maybe a locally run model on my GPU at the moment.", "context": "maintains ~20 projects", "use_case": "self-hosted related coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 795685536, "item_id": 49730346, "url": "https://news.ycombinator.com/item?id=49730346", "by": "pimeys", "date": "2026-09-16", "stance": "recommend", "quote": "V4.1 is really very close to SOTA in my books", "excerpt": "My experience is complete opposite from yours. The V4 Flash was already quite good, but V4.1 is really very close to SOTA in my books. I've eval'd these models for weeks against Gemini 3.8, Kimi K3 and Opus 5, and DeepSeek absolutely wins these evals. It's really crazy. We price per token our customers. If Kimi was about 30% of the price of Opus 5 for the same quality, DeepSeek is 1/10th of a price of Kimi K3. We've ", "context": "prices per token to customers, runs 7-8 parallel agents", "use_case": "agentic coding work", "switched_from": "Kimi K3", "story": "DeepSeek v4.1 Flash Is Now Our Best Hacking Model"}]}, {"label": "Speed", "summary": "Users report roughly 150 to 200 tokens per second and call it faster than Opus or GPT models.", "quotes": [{"id": 802516448, "item_id": 50157278, "url": "https://news.ycombinator.com/item?id=50157278", "by": "sinuhe69", "date": "2026-10-19", "stance": "recommend", "quote": "the DeepSeek 4.1 models are currently the ultimate cost-effective model the rest have to match with", "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": "cost-effective coding", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 800205984, "item_id": 50012874, "url": "https://news.ycombinator.com/item?id=50012874", "by": "aussieguy1234", "date": "2026-10-08", "stance": "recommend", "quote": "What blows me away about this model is it's speed.", "excerpt": "What blows me away about this model is it's speed. It's way faster than Opus or any of the GPT models. I have a coding harness which is opencode plus a few skills relevant to my workflow. Deepseek 4.1 Flash does very well in this environment. I haven't noticed much difference quality wise compared to Opus 5, which I use in my day job as my employer pays for it (although I'm considering using DeepSeek here too given h", "context": "", "use_case": "coding with opencode", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 804179970, "item_id": 50261248, "url": "https://news.ycombinator.com/item?id=50261248", "by": "Roark66", "date": "2026-10-27", "stance": "recommend", "quote": "you can spin a bunch of DeepSeek Flash agents and each will run at ~150-200t/s", "excerpt": "Actually more AI agents do the work faster and cheaper. I use such orchestration for those 2 reasons. One, a supervisor model like Opus 5.5 (or fable before they lobotomised it) limited to 200k of context and told to keep it's progress updated in a text file is smart enough to understand the whole task. It farms smaller sub tasks to a fleet of 7 Qwen3.8-Flash-Next agents. Not only it uses up 10% of the tokens opus wo", "context": "", "use_case": "sub-task agents", "switched_from": "", "story": "The Mythical Man-Month in the Age of AI"}]}, {"label": "Coding ability", "summary": "Handles daily work in Python, C++, CUDA, Rust and Svelte, and some put it close to state-of-the-art models.", "quotes": [{"id": 803966624, "item_id": 50247914, "url": "https://news.ycombinator.com/item?id=50247914", "by": "joshheitzman", "date": "2026-10-26", "stance": "recommend", "quote": "I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "excerpt": "As a counter-example, I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "context": "", "use_case": "daily coding on multiple projects", "switched_from": "", "story": "Anthropic S-1"}, {"id": 795764944, "item_id": 49735309, "url": "https://news.ycombinator.com/item?id=49735309", "by": "rapind", "date": "2026-09-17", "stance": "recommend", "quote": "I\u2019ve been very pleased with DS 4.1 flash.", "excerpt": "I\u2019ve been very pleased with DS 4.1 flash. Not so much the 4.0 models, but for coding (Rust) it\u2019s been great so far (3 solid days of work). I\u2019ll give Mimo a try.", "context": "", "use_case": "Rust coding", "switched_from": "", "story": "Xiaomi Mimo 2.6 live post-training dashboard"}, {"id": 798403264, "item_id": 49900204, "url": "https://news.ycombinator.com/item?id=49900204", "by": "thiht", "date": "2026-09-29", "stance": "recommend", "quote": "I've come to forget I was not using Claude, it's a really good model", "excerpt": "I've been using Claude Code at work and OpenCode for side projects for a few months. Every OpenCode model I've tried always felt subpar compared to Claude, but good enough. But it changed with DeepSeek 4.1 Flash, I've been using it for the past few days and I've come to forget I was not using Claude, it's a really good model and it's basically free for my usage (I used it almost all the weekend and spent ~$5)", "context": "", "use_case": "side project coding", "switched_from": "", "story": "GPT 6.1 Sol: Near-Astra intelligence for a fifth of the price"}]}, {"label": "Implementation workhorse", "summary": "Frequently paired with a frontier planner model, with Flash doing the implementation and sub-agent work.", "quotes": [{"id": 795660832, "item_id": 49728802, "url": "https://news.ycombinator.com/item?id=49728802", "by": "surgical_fire", "date": "2026-09-16", "stance": "recommend", "quote": "I love DS flash, it is an amazing workhorse to implement plans created by more robust models", "excerpt": "What I find surprising is that DS Flash can do it at all. I love DS flash, it is an amazing workhorse to implement plans created by more robust models (such as GLM). But a more fair comparison would be of DS Flash with GLM Flash.", "context": "", "use_case": "implementing plans", "switched_from": "", "story": "DeepSeek v4.1 Flash Is Now Our Best Hacking Model"}, {"id": 800250993, "item_id": 50015687, "url": "https://news.ycombinator.com/item?id=50015687", "by": "yuhmahp", "date": "2026-10-09", "stance": "recommend", "quote": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "excerpt": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "context": "", "use_case": "general coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 802763590, "item_id": 50172724, "url": "https://news.ycombinator.com/item?id=50172724", "by": "CharlieDigital", "date": "2026-10-20", "stance": "recommend", "quote": "Coding is handled by Haiku 5.5/GPT 6 Luna/DeepSeek v4.1 Flash.", "excerpt": "I use the frontier models only in plan mode; coding is held out to cheaper models. I run Paseo as an orchestrator with agent profiles. \"Distinguished Member of Technical Staff\" -> Fable, plan mode. \"End-User Customer Advocate\" -> Astra, plan mode. A few others. Basically, these agents (with profile prompts) can only be \"consulted\" for input on design elements; it's OK if they produce a bit of verbosity here. This has", "context": "", "use_case": "coding", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}]}, {"label": "Debugging", "summary": "Credited with quickly finding bugs, including in difficult brownfield code.", "quotes": [{"id": 800358800, "item_id": 50022425, "url": "https://news.ycombinator.com/item?id=50022425", "by": "erans", "date": "2026-10-09", "stance": "recommend", "quote": "I've been using DS 4.1 Flash for a while now as my main workhorse. It's amazing.", "excerpt": "I've been using DS 4.1 Flash for a while now as my main workhorse. It's amazing. It's fast and it is also a GREAT debugger - it might not always find the best solution but it will definitely find the problem quickly. I'm also the CEO of a new company - LunaRoute - so if you want private, fixed cost, we run our own server in the US type of DeepSeek 4.1 Flash check out https://www.lunaroute.com If you want a trial, hit", "context": "CEO of LunaRoute", "use_case": "main workhorse and debugging", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 804015808, "item_id": 50250988, "url": "https://news.ycombinator.com/item?id=50250988", "by": "wren6991", "date": "2026-10-26", "stance": "recommend", "quote": "V4.1 Flash 1017 is my daily driver cloud model.", "excerpt": "Yes, V4.1 Flash 1017 is my daily driver cloud model. Copes fine with what I'd consider difficult brownfield development, like fixing bugs in vLLM prompt cache and coming up with small/sane overlay patches for them. Maybe it's disappointing for people who want to one-shot a game or website from a single prompt; I haven't tried that. It's nice that their peak-rate hours end quite early in the day UK time.", "context": "", "use_case": "brownfield development, bug fixing", "switched_from": "", "story": "GPT-6.1 Terra"}]}, {"label": "Local running", "summary": "Some run quantized versions locally, such as on a 128GB Mac Studio, with decent coherence.", "quotes": [{"id": 802380992, "item_id": 50148812, "url": "https://news.ycombinator.com/item?id=50148812", "by": "davedx", "date": "2026-10-19", "stance": "recommend", "quote": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability.", "excerpt": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability. Apparently the V4.1 model can also be run by streaming weights from the SSD - I've not tried this yet because it results in a lower TPS, but eventually this version of 4.1 should also be usable like this.", "context": "runs on Mac Studio with 128GB RAM", "use_case": "local coding model", "switched_from": "", "story": "DeepSeek V4.1 Flash 1017 (Max)"}, {"id": 799810305, "item_id": 49988144, "url": "https://news.ycombinator.com/item?id=49988144", "by": "colingauvin", "date": "2026-10-07", "stance": "recommend", "quote": "Then DeepSeek 4 Flash came out and suddenly it was like... Ok. You can run a local model that keeps coherence, and is decent technically.", "excerpt": "The author's configuration is..... Interesting. I don't disagree with the title. This spring folks were saying \"I use Qwen 3.6 27B (or 35 A3B) as a daily driver\". They either had low standards or were lying. The models were good enough technically, they capable enough, that you could see the potential, but the coherence was really not great compared to frontier. They'd lose track of what they were doing, mix things u", "context": "runs models on 2x Sparks", "use_case": "local model", "switched_from": "", "story": "Local AI is real now. And it's blowing my mind"}, {"id": 801427744, "item_id": 50089234, "url": "https://news.ycombinator.com/item?id=50089234", "by": "ryan_glass", "date": "2026-10-14", "stance": "recommend", "quote": "I use Deepseek V4 flash 0731 Q8 quantisation from unsolth with reasoning turned off on a local server for quick easy tasks", "excerpt": "I thought it would be interesting to see if the local AI I use for quick question and answer stuff could crack the code. I use Deepseek V4 flash 0731 Q8 quantisation from unsolth with reasoning turned off on a local server for quick easy tasks. Note I use another model locally for harder tasks that need reasoning. I was a bit dissappointed that it knew it was Poe's cypher so didn't need to even try to crack the code.", "context": "", "use_case": "local quick Q&A tasks", "switched_from": "", "story": "The Gold-Bug: Edgar Allan Poe's story that taught me about cryptography (2024)"}]}], "caveats": [{"label": "Below frontier", "summary": "Some report weaker solutions than Opus or Sol, rabbit holes, false assertions, or overall disappointing results.", "quotes": [{"id": 800204064, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "orchestrator, tester, subagent", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800403126, "item_id": 50025195, "url": "https://news.ycombinator.com/item?id=50025195", "by": "aftbit", "date": "2026-10-09", "stance": "warn", "quote": "they're more likely to go down rabbit holes, confidently assert falsehoods, or fix bugs with changes to my test harness rather than my code.", "excerpt": "I have had good luck rotating between the three tiers of GPT 5.6 with occasional jumps up to Astra or Fable. Most of my work is with GPT-5.6-Sol. Simple tasks like data extraction or trivial refactors (rename this variable etc) I often push down to Terra or Luna, or even to self-hosted Gemma4:31B. Very tricky stuff, like planning a new feature, design review, or code review of a complex change across multiple reposit", "context": "", "use_case": "easy coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800234272, "item_id": 50014642, "url": "https://news.ycombinator.com/item?id=50014642", "by": "ltbarcly3", "date": "2026-10-09", "stance": "warn", "quote": "DeepSeek 4.1 Flash kindof sucks.", "excerpt": "DeepSeek 4.1 Flash kindof sucks. I used it a bunch and it kindof sucks. I don't know if they are gaming benchmarks or what. Luna is on par in benchmarks and my personal experience is Luna is better for what I do, and Luna is cheaper. Comparing Deepseek 4.1 flash to Opus is just ludicrous. https://artificialanalysis.ai/models/releases/comparisons?compare=gpt-6-luna%2Cdeepseek-v4-1-flash", "context": "", "use_case": "", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}]}, {"label": "Token verbosity", "summary": "It can be chatty, using about 3x the tokens of Sol per task, so cost per task matters more than token price.", "quotes": [{"id": 800198496, "item_id": 50012406, "url": "https://news.ycombinator.com/item?id=50012406", "by": "rapind", "date": "2026-10-08", "stance": "recommend", "quote": "I do think it's the most viable alternative I've seen so far", "excerpt": "It's really not cheaper than frontier subscriptions. It's getting closer, and it's a great model, but it is not more value per task than the frontier subscriptions. Don't be swayed by the token costs, it's very chatty, like 3x more tokens for the same task as sol. I used dsf 4.1 full time for about a week. It blows frontier API pricing out of the water, but again, look at cost per task, not token usage. Still easily ", "context": "", "use_case": "full-time coding work", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 798302640, "item_id": 49893915, "url": "https://news.ycombinator.com/item?id=49893915", "by": "dools", "date": "2026-09-29", "stance": "recommend", "quote": "Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do.", "excerpt": "I go for weeks without topping up my Deepseek balance, and have to top up my OpenAI balance at least every week. GPT 6 Luna closed the gap significantly for sure (it seems to be about twice as expensive as DS v4.1 Flash), but Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do. Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt, ", "context": "", "use_case": "planning and implementation", "switched_from": "", "story": "Sonnet 5.5"}]}, {"label": "Poor explanations", "summary": "Described as bad at grilling sessions and explaining technical decisions, and annoying to chat with in some roles.", "quotes": [{"id": 800204064, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "orchestrator, tester, subagent", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 802549169, "item_id": 50159323, "url": "https://news.ycombinator.com/item?id=50159323", "by": "bitexploder", "date": "2026-10-19", "stance": "recommend", "quote": "I really like DS Flash too, it can just be /really/ annoying in some roles", "excerpt": "I know I mention it way too often right now, but I am just so impressed with Qwen Flash Next right now. I have been using it locally and it's just such a great little model to work with. It makes mistakes and has its quirks, but in general, it just gets things done and its default manner is just right. It is not overly proactive or making messes for me, even compared to DeepSeek Flash 4.1. I really like DS Flash too,", "context": "", "use_case": "chat and coding", "switched_from": "", "story": "MiniMax M3.1 Flash"}]}, {"label": "Provider variance", "summary": "Inference quality and configuration reportedly vary between hosting providers and quantizations.", "quotes": [{"id": 803449648, "item_id": 50215603, "url": "https://news.ycombinator.com/item?id=50215603", "by": "joshheitzman", "date": "2026-10-23", "stance": "recommend", "quote": "I recommend trying deepseek v4.1 flash from together.ai", "excerpt": "I use that specific model as my primary LLM for coding and diagnostics on projects in Python, C++, CUDA, and for web research. I'm also using it with my own harness (you mentioned \"in my harness\"), so I'll share two things I found about using open weight model providers. The first is that getting the configuration right isn't a per model thing, its a per model per provider per point in time thing. The point in time t", "context": "uses a self-built harness for coding in Python, C++, CUDA", "use_case": "coding with open-weight model", "switched_from": "", "story": "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"}, {"id": 802698145, "item_id": 50168634, "url": "https://news.ycombinator.com/item?id=50168634", "by": "mixermachine", "date": "2026-10-20", "stance": "recommend", "quote": "The value you get is claimed to be 80$ in DeepSeek v4.1 Flash API value.", "excerpt": "You could consider going with OpenCode Go for 10$. The value you get is claimed to be 80$ in DeepSeek v4.1 Flash API value. I use that and so far it is good for me. Some claim they run a different quant but I so far don't see that.", "context": "", "use_case": "coding via OpenCode Go", "switched_from": "", "story": "DeepInfer: High-performance inference framework for DeepSeek models"}]}, {"label": "Hardware demands", "summary": "Running it locally at good quality needs expensive hardware.", "quotes": [{"id": 800546449, "item_id": 50034153, "url": "https://news.ycombinator.com/item?id=50034153", "by": "AustinDev", "date": "2026-10-10", "stance": "recommend", "quote": "Running DeepSeek 4.1 Flash at FP4 is going to need at least a 10-15k machine. I'd consider that near frontier level.", "excerpt": ">You can already performantly run frontier level models on systems with $1600 tier video cards. Quanted out Qwen 3.8 models are not frontier level. Have you used them for real work? They suck. They are so unsure of themselves due to the quant-lobotomy. They just endlessly loop in thinking even when they have the right decision and then output it when they get to the thinking cap. They score close in benchmarks becaus", "context": "", "use_case": "", "switched_from": "", "story": "100+ reactions to 100+ solutions"}, {"id": 802380992, "item_id": 50148812, "url": "https://news.ycombinator.com/item?id=50148812", "by": "davedx", "date": "2026-10-19", "stance": "recommend", "quote": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability.", "excerpt": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability. Apparently the V4.1 model can also be run by streaming weights from the SSD - I've not tried this yet because it results in a lower TPS, but eventually this version of 4.1 should also be usable like this.", "context": "runs on Mac Studio with 128GB RAM", "use_case": "local coding model", "switched_from": "", "story": "DeepSeek V4.1 Flash 1017 (Max)"}]}], "who": [{"by": "d2kx", "context": "EU-based company, B2B2C app", "url": "https://news.ycombinator.com/item?id=50290261", "date": "2026-10-29"}, {"by": "potsandpans", "context": "decompiling a PlayStation game", "url": "https://news.ycombinator.com/item?id=50222286", "date": "2026-10-24"}, {"by": "joshheitzman", "context": "uses a self-built harness for coding in Python, C++, CUDA", "url": "https://news.ycombinator.com/item?id=50215603", "date": "2026-10-23"}, {"by": "davedx", "context": "runs on Mac Studio with 128GB RAM", "url": "https://news.ycombinator.com/item?id=50148812", "date": "2026-10-19"}, {"by": "wereHamster", "context": "maintains a product receiving vulnerability reports", "url": "https://news.ycombinator.com/item?id=50100730", "date": "2026-10-15"}, {"by": "erans", "context": "CEO of LunaRoute", "url": "https://news.ycombinator.com/item?id=50022425", "date": "2026-10-09"}, {"by": "Primer81", "context": "maintains ~20 projects", "url": "https://news.ycombinator.com/item?id=50019691", "date": "2026-10-09"}, {"by": "georgel", "context": "uses it 5-8 hrs a day for coding", "url": "https://news.ycombinator.com/item?id=50012838", "date": "2026-10-08"}, {"by": "lmf4lol", "context": "runs personal assistants for their team", "url": "https://news.ycombinator.com/item?id=50011347", "date": "2026-10-08"}, {"by": "colingauvin", "context": "runs models on 2x Sparks", "url": "https://news.ycombinator.com/item?id=49988144", "date": "2026-10-07"}, {"by": "sieve", "context": "validates behavior instead of reading code", "url": "https://news.ycombinator.com/item?id=49811750", "date": "2026-09-23"}, {"by": "mobelkh", "context": "uses open and closed models since Sonnet 4", "url": "https://news.ycombinator.com/item?id=49797780", "date": "2026-09-22"}], "tags": ["for coding agents", "for multi-agent orchestration", "self-hostable", "for budget-conscious devs", "for side projects"], "use_cases": ["coding", "coding tasks", "local inference", "coding via opencode", "coding agent", "local coding model"], "compared": [{"slug": "claude-opus", "name": "Claude Opus", "co": 46, "switched_from": 2, "switched_to": 0}, {"slug": "glm", "name": "GLM", "co": 34, "switched_from": 2, "switched_to": 0}, {"slug": "opencode", "name": "OpenCode", "co": 36, "switched_from": 0, "switched_to": 0}, {"slug": "claude", "name": "Claude", "co": 22, "switched_from": 6, "switched_to": 0}, {"slug": "kimi", "name": "Kimi", "co": 20, "switched_from": 2, "switched_to": 1}, {"slug": "deepseek", "name": "DeepSeek", "co": 24, "switched_from": 1, "switched_to": 0}, {"slug": "pi", "name": "Pi", "co": 25, "switched_from": 0, "switched_to": 0}, {"slug": "glm-flash", "name": "GLM Flash", "co": 20, "switched_from": 1, "switched_to": 1}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=50000488", "title": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?", "mentions": 20}, {"url": "https://news.ycombinator.com/item?id=49672966", "title": "Ask HN: What default model do you use and why?", "mentions": 14}, {"url": "https://news.ycombinator.com/item?id=49214008", "title": "DeepSeek V4 Flash 0731", "mentions": 11}, {"url": "https://news.ycombinator.com/item?id=49119559", "title": "DeepSeek-V4-Flash Update", "mentions": 7}, {"url": "https://news.ycombinator.com/item?id=50129045", "title": "DeepSeek-v4.1-Flash-1017", "mentions": 6}, {"url": "https://news.ycombinator.com/item?id=48542100", "title": "Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding?", "mentions": 6}], "sentiment_by_month": {"2025-11": {}, "2025-12": {}, "2026-01": {}, "2026-02": {}, "2026-03": {}, "2026-04": {"recommend": 1}, "2026-05": {"neutral": 2, "recommend": 6}, "2026-06": {"recommend": 28, "compare": 1, "neutral": 3, "warn": 1}, "2026-07": {"recommend": 19, "neutral": 7, "warn": 1}, "2026-08": {"recommend": 45, "neutral": 7, "compare": 4}, "2026-09": {"neutral": 11, "recommend": 33, "warn": 1, "compare": 4}, "2026-10": {"recommend": 39, "neutral": 8, "compare": 3, "warn": 2}}, "quotes": [{"id": 804644178, "item_id": 50290261, "url": "https://news.ycombinator.com/item?id=50290261", "by": "d2kx", "date": "2026-10-29", "stance": "neutral", "quote": "We use Mistral Large 4, GLM 5.3 Flash and DeepSeek v4.1 Flash 1017 through their API", "excerpt": "Mistral has been great for us as a provider for our EU-based company. We use Mistral Large 4, GLM 5.3 Flash and DeepSeek v4.1 Flash 1017 through their API for our B2B2C app for complex work around building blocks for documents and custom tools. A few months ago we weren't sure which EU provider to use, but since they introduced 3rd party models, added more features (now Responses API etc.) and improved their own mode", "context": "EU-based company, B2B2C app", "use_case": "document building blocks, custom tools", "switched_from": "", "story": "We're releasing the weights of Mistral Large 4"}, {"id": 804330976, "item_id": 50270686, "url": "https://news.ycombinator.com/item?id=50270686", "by": "piterrro", "date": "2026-10-28", "stance": "recommend", "quote": "I'm still using DS-flash - it beats frontier model on speed and cost keeping good enough code quality.", "excerpt": "I'm still using DS-flash - it beats frontier model on speed and cost keeping good enough code quality. It depends on the harness and constraints you give to the model but for most of the tasks its a great fit, I have not switched from DS since couple of weeks already (previously I would switch to Opus for some heavier work but I stopped recently). My harness consists of multi-level AGENTS.md and SOUL.md containing \"g", "context": "", "use_case": "coding tasks", "switched_from": "Opus", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 804330770, "item_id": 50270673, "url": "https://news.ycombinator.com/item?id=50270673", "by": "wyrdcurt", "date": "2026-10-28", "stance": "recommend", "quote": "Deepseek 4.1 Flash is turning out to be a sweet spot for me; it's cheap and blazing fast", "excerpt": "I've personally experienced overthinking and the thinking loop issue more often than seeing it not think enough, I know what you mean though. I tried Haiku 5.5 and the most impressive thing about it to me was that it seemed not to overthink when it wasn't necessary, yet it somehow knew when it should spend some more time reasoning about a task. American models are too rich for my blood though. I use GLM 5.3 Flash mor", "context": "", "use_case": "running agents", "switched_from": "", "story": "Artificial Analysis Intelligence Index v5"}, {"id": 804214530, "item_id": 50263408, "url": "https://news.ycombinator.com/item?id=50263408", "by": "colingauvin", "date": "2026-10-27", "stance": "neutral", "quote": "right now that's DeepSeek 4.1 Flash 1017", "excerpt": "One thing I have been using LFM 2.6 for is annotating emails, text messages, and other metadata. I keep it all ingesting into one giant database and annotate it all as it comes in on an RTX A4500. The LFM annotations are basically triage, and then slowly annotations trickle into my halo model based on lowest confidence (right now that's DeepSeek 4.1 Flash 1017) for an upgraded annotation. (Agents can then search this", "context": "", "use_case": "upgraded annotation", "switched_from": "", "story": "Introducing LFM3: Every Modality for Specialized Intelligence"}, {"id": 804179970, "item_id": 50261248, "url": "https://news.ycombinator.com/item?id=50261248", "by": "Roark66", "date": "2026-10-27", "stance": "recommend", "quote": "you can spin a bunch of DeepSeek Flash agents and each will run at ~150-200t/s", "excerpt": "Actually more AI agents do the work faster and cheaper. I use such orchestration for those 2 reasons. One, a supervisor model like Opus 5.5 (or fable before they lobotomised it) limited to 200k of context and told to keep it's progress updated in a text file is smart enough to understand the whole task. It farms smaller sub tasks to a fleet of 7 Qwen3.8-Flash-Next agents. Not only it uses up 10% of the tokens opus wo", "context": "", "use_case": "sub-task agents", "switched_from": "", "story": "The Mythical Man-Month in the Age of AI"}, {"id": 804024545, "item_id": 50251534, "url": "https://news.ycombinator.com/item?id=50251534", "by": "l3x4ur1n", "date": "2026-10-26", "stance": "neutral", "quote": "I use DeepSeek flash API as a driver most of the time.", "excerpt": "I am using Hermes and I did not try charts, but the rest it can do with ease. And I use DeepSeek flash API as a driver most of the time.", "context": "", "use_case": "driver model for agent", "switched_from": "", "story": "Show HN: Hob - Open-source, on device alternative to Claude Cowork and ChatGPT"}, {"id": 804015808, "item_id": 50250988, "url": "https://news.ycombinator.com/item?id=50250988", "by": "wren6991", "date": "2026-10-26", "stance": "recommend", "quote": "V4.1 Flash 1017 is my daily driver cloud model.", "excerpt": "Yes, V4.1 Flash 1017 is my daily driver cloud model. Copes fine with what I'd consider difficult brownfield development, like fixing bugs in vLLM prompt cache and coming up with small/sane overlay patches for them. Maybe it's disappointing for people who want to one-shot a game or website from a single prompt; I haven't tried that. It's nice that their peak-rate hours end quite early in the day UK time.", "context": "", "use_case": "brownfield development, bug fixing", "switched_from": "", "story": "GPT-6.1 Terra"}, {"id": 803966624, "item_id": 50247914, "url": "https://news.ycombinator.com/item?id=50247914", "by": "joshheitzman", "date": "2026-10-26", "stance": "recommend", "quote": "I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "excerpt": "As a counter-example, I use Deepseek v4.1 Flash to make progress on multiple projects (Python, C++, CUDA) every single day.", "context": "", "use_case": "daily coding on multiple projects", "switched_from": "", "story": "Anthropic S-1"}, {"id": 803556579, "item_id": 50222286, "url": "https://news.ycombinator.com/item?id=50222286", "by": "potsandpans", "date": "2026-10-24", "stance": "recommend", "quote": "deepseek flash variants, work horse.", "excerpt": "I recently just crossed the threshold for 100% decompilation of a PlayStation game. I used a lot of models in the process, and I feel like this is a good demonstration the \"sauce\" that opus has. The harness I've built for it is pretty straight forward: split binary into individual functions, give the agent a bunch of small tools to diff output with underlying binary, make them wrote clean c until there's a certified ", "context": "decompiling a PlayStation game", "use_case": "decompiling moderately complex functions", "switched_from": "", "story": "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"}, {"id": 803494437, "item_id": 50218402, "url": "https://news.ycombinator.com/item?id=50218402", "by": "sail0rm00n", "date": "2026-10-24", "stance": "recommend", "quote": "Gemini Flash 3.8, DeepSeek Flash, and Luna are still up there as favorites", "excerpt": "As tribute, I bit the bullet on the $40 plan to test out K3.1 Pro. Still burning through the 5-hour rate limit way too quickly, though. Output quality seems solid enough, but the value will probably make a lot more sense via API pricing for smaller projects. My current tier list based on what I\u2019ve been testing: 1. Opus 5.5 2. Astra 3. Argon 4. Kimi K3.1 Pro (For lightweight/daily driver tasks, Gemini Flash 3.8, DeepS", "context": "", "use_case": "lightweight daily tasks", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 803457888, "item_id": 50216118, "url": "https://news.ycombinator.com/item?id=50216118", "by": "LeFantome", "date": "2026-10-23", "stance": "recommend", "quote": "I have been using DS Flash 4.1 pretty heavily. It runs for hours and hours and is pretty cheap.", "excerpt": "I have been using DS Flash 4.1 pretty heavily. It runs for hours and hours and is pretty cheap. I realized the other day that I had a $30 a month Claude subscription. So, I fired it up all excited and tried Opus 5.5 on a codebase I was already in with DS. After about 30 minutes I hit a 5 hour limit. The next day I tried again and the same thing happened. And it felt much slower to boot. I was super disappointed.", "context": "", "use_case": "coding on a codebase", "switched_from": "", "story": "Kimi K3.1 Pro and K3.1 Pro Highspeed"}, {"id": 803449648, "item_id": 50215603, "url": "https://news.ycombinator.com/item?id=50215603", "by": "joshheitzman", "date": "2026-10-23", "stance": "recommend", "quote": "I recommend trying deepseek v4.1 flash from together.ai", "excerpt": "I use that specific model as my primary LLM for coding and diagnostics on projects in Python, C++, CUDA, and for web research. I'm also using it with my own harness (you mentioned \"in my harness\"), so I'll share two things I found about using open weight model providers. The first is that getting the configuration right isn't a per model thing, its a per model per provider per point in time thing. The point in time t", "context": "uses a self-built harness for coding in Python, C++, CUDA", "use_case": "coding with open-weight model", "switched_from": "", "story": "Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line"}, {"id": 803314626, "item_id": 50207164, "url": "https://news.ycombinator.com/item?id=50207164", "by": "phillc73", "date": "2026-10-23", "stance": "neutral", "quote": "Deepseek-4.1 Flash 1017", "excerpt": "I use Maki[1] harness and GML-5.3, GLM5.3 Flash and Deepseek-4.1 Flash 1017, with a little bit of MiMo-2.6. The key I've found is a strong skill. At the beginning, sure there were problems. However, I implemented an AGENTS rule that said if a Janet script failed more than twice on the same \"gotcha\", the correct syntax was added to the skill. The skill is regularly reviewed, consolidated and pruned. The Janet skill is", "context": "", "use_case": "coding with Janet", "switched_from": "", "story": "Jank \u2013 The native Clojure dialect with seamless C++ interop"}, {"id": 803033600, "item_id": 50189600, "url": "https://news.ycombinator.com/item?id=50189600", "by": "brazukadev", "date": "2026-10-22", "stance": "recommend", "quote": "It is terribly fast! And smart. And cheap.", "excerpt": "I used to think that until I started using DeepSeek V4.1 Flash 1017 for the past 2 days (what a bad name). It is terribly fast! And smart. And cheap.", "context": "", "use_case": "", "switched_from": "", "story": "Token-maxxing is dead, OpenRouter's token usage plateaus"}, {"id": 802763590, "item_id": 50172724, "url": "https://news.ycombinator.com/item?id=50172724", "by": "CharlieDigital", "date": "2026-10-20", "stance": "recommend", "quote": "Coding is handled by Haiku 5.5/GPT 6 Luna/DeepSeek v4.1 Flash.", "excerpt": "I use the frontier models only in plan mode; coding is held out to cheaper models. I run Paseo as an orchestrator with agent profiles. \"Distinguished Member of Technical Staff\" -> Fable, plan mode. \"End-User Customer Advocate\" -> Astra, plan mode. A few others. Basically, these agents (with profile prompts) can only be \"consulted\" for input on design elements; it's OK if they produce a bit of verbosity here. This has", "context": "", "use_case": "coding", "switched_from": "", "story": "SlopCodeBench 10/2026 \u2013 Opus 5.5, Astra wins overall, xhigh considered harmful"}, {"id": 802698145, "item_id": 50168634, "url": "https://news.ycombinator.com/item?id=50168634", "by": "mixermachine", "date": "2026-10-20", "stance": "recommend", "quote": "The value you get is claimed to be 80$ in DeepSeek v4.1 Flash API value.", "excerpt": "You could consider going with OpenCode Go for 10$. The value you get is claimed to be 80$ in DeepSeek v4.1 Flash API value. I use that and so far it is good for me. Some claim they run a different quant but I so far don't see that.", "context": "", "use_case": "coding via OpenCode Go", "switched_from": "", "story": "DeepInfer: High-performance inference framework for DeepSeek models"}, {"id": 802679059, "item_id": 50167441, "url": "https://news.ycombinator.com/item?id=50167441", "by": "A_D_E_P_T", "date": "2026-10-20", "stance": "recommend", "quote": "Deepseek 4.1 Flash in OpenCode.", "excerpt": "The Google models are great \"general purpose\" and research LLMs, but they're relatively poor at coding. The way I see it, you have two options. 1) Get on the Claude Max 20x plan and use Opus 5.5 for review/orchestration and Haiku 5.5 for grunt work. Haiku 5.5 is decent, cheap, and fast. 2) Deepseek 4.1 Flash in OpenCode. You should also test MiMo v2.6. It's a little bit uneven, but might work well for whatever you're", "context": "", "use_case": "cheap AI coding", "switched_from": "", "story": "Ask HN: Best value/money AI coding plan bundle?"}, {"id": 802562640, "item_id": 50160165, "url": "https://news.ycombinator.com/item?id=50160165", "by": "seabombs", "date": "2026-10-20", "stance": "neutral", "quote": "I use DS-4.1-Flash a lot but find it often gets caught up in long thinking runs.", "excerpt": "I use DS-4.1-Flash a lot but find it often gets caught up in long thinking runs. The \"But wait... Actually... Hmm...\" type of thing. Even though the tok/s is good, it stills feels slow because of this. Same with 1017. I use Opus 5.5 at work and it does this much less (I know, different classes). Maybe it is to be expected of the flash models.", "context": "", "use_case": "coding with AI", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802549169, "item_id": 50159323, "url": "https://news.ycombinator.com/item?id=50159323", "by": "bitexploder", "date": "2026-10-19", "stance": "recommend", "quote": "I really like DS Flash too, it can just be /really/ annoying in some roles", "excerpt": "I know I mention it way too often right now, but I am just so impressed with Qwen Flash Next right now. I have been using it locally and it's just such a great little model to work with. It makes mistakes and has its quirks, but in general, it just gets things done and its default manner is just right. It is not overly proactive or making messes for me, even compared to DeepSeek Flash 4.1. I really like DS Flash too,", "context": "", "use_case": "chat and coding", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802518400, "item_id": 50157400, "url": "https://news.ycombinator.com/item?id=50157400", "by": "RussianCow", "date": "2026-10-19", "stance": "recommend", "quote": "I have a $20/month OpenAI subscription that I supplement with DeepSeek V4.1 Flash", "excerpt": "> Claude sub has a 60x subsidy, so a $20 sub gives you $1200 in API. This is only true if you max out the plan, which requires working around awkward 5-hour and weekly limits. For side projects, sure, you can use a Ralph loop or whatever and just continue when the window resets. For work, I can't afford to simply stop working when I hit my limit. And using Claude at API prices is prohibitively expensive. Personally, ", "context": "", "use_case": "primary coding model", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802517969, "item_id": 50157373, "url": "https://news.ycombinator.com/item?id=50157373", "by": "oooodo", "date": "2026-10-19", "stance": "recommend", "quote": "not as concise and direct as DeepSeek 4.1 Flash in coding", "excerpt": "I used it as Space Bunny, was very good at documentation, structuring and organizing the things, but was not as concise and direct as DeepSeek 4.1 Flash in coding, we could not focus on code part as a human and AI team from how many approves are needed for each step discussion.", "context": "", "use_case": "coding", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802516448, "item_id": 50157278, "url": "https://news.ycombinator.com/item?id=50157278", "by": "sinuhe69", "date": "2026-10-19", "stance": "recommend", "quote": "the DeepSeek 4.1 models are currently the ultimate cost-effective model the rest have to match with", "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": "cost-effective coding", "switched_from": "", "story": "MiniMax M3.1 Flash"}, {"id": 802501105, "item_id": 50156319, "url": "https://news.ycombinator.com/item?id=50156319", "by": "segmondy", "date": "2026-10-19", "stance": "recommend", "quote": "Both were great", "excerpt": "Too early to call, I ran it through the weekend and compared with Qwen3.8-Flash-Next, DSv41F-1017 at FP8 and Q3.8FN at Q8. Both were great, but surprisingly, Q3.8FN had the better result. I was modelling and generating simulation code for a complex system. I used both because I had a bunch of images that needed to be used for input to steer the output. I think these models have strengths and it's best to experiment a", "context": "", "use_case": "simulation code generation", "switched_from": "", "story": "DeepSeek V4.1 Flash 1017 (Max)"}, {"id": 802380992, "item_id": 50148812, "url": "https://news.ycombinator.com/item?id=50148812", "by": "davedx", "date": "2026-10-19", "stance": "recommend", "quote": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability.", "excerpt": "I use DwarfStar DeekSeek V4 Flash on my Mac Studio with 128gb of RAM and it's pretty good - decent TPS and coding ability. Apparently the V4.1 model can also be run by streaming weights from the SSD - I've not tried this yet because it results in a lower TPS, but eventually this version of 4.1 should also be usable like this.", "context": "runs on Mac Studio with 128GB RAM", "use_case": "local coding model", "switched_from": "", "story": "DeepSeek V4.1 Flash 1017 (Max)"}, {"id": 802219520, "item_id": 50138720, "url": "https://news.ycombinator.com/item?id=50138720", "by": "dsco", "date": "2026-10-18", "stance": "recommend", "quote": "I use this thing daily and it\u2019s neck and neck with SOTA models.", "excerpt": "DeepSeek Flash really came at a bad time before the Anthropic IPO. I use this thing daily and it\u2019s neck and neck with SOTA models. The search is one of the things setting it apart.", "context": "", "use_case": "daily AI use with search", "switched_from": "", "story": "DeepSeek search and fetch \u2013 $0.75 per 1000 successful requests"}, {"id": 802192642, "item_id": 50137040, "url": "https://news.ycombinator.com/item?id=50137040", "by": "markasoftware", "date": "2026-10-18", "stance": "compare", "quote": "is quite smarter than dsv4 flash", "excerpt": "Fwiw I find gpt Sol on medium or low effort to also be faster than opus to the point of helping me not check out, but is quite smarter than dsv4 flash", "context": "", "use_case": "coding", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802095810, "item_id": 50130988, "url": "https://news.ycombinator.com/item?id=50130988", "by": "phillc73", "date": "2026-10-17", "stance": "recommend", "quote": "Quite happily chugging along with a nice balance across them now.", "excerpt": "I\u2019ll add Melious.[1] I feel like I\u2019m getting good value out of my subscription and they have the cheapest GLM-5.3 API rates I could find in Europe. I moved to them recently from Mistral for the better prices. They do have a slightly different daily inference capacity allowance, but it\u2019s actually not too bad once used to it. I was forced to up my game and figure out what GLM-5.3 Flash and Deepseek-4.1 Flash could actu", "context": "", "use_case": "general LLM tasks", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802090929, "item_id": 50130683, "url": "https://news.ycombinator.com/item?id=50130683", "by": "drob518", "date": "2026-10-17", "stance": "neutral", "quote": "DS 4 Flash, my previous daily driver, which I also found quite good", "excerpt": "Interesting. I\u2019ve been using GLM 5.3 Flash for a couple months and found it to be great. What has it been doing that you found to be \u201ctoo clever\u201d? It sometimes gets stuck in loops. I once asked it to do a code review of a simple pattern and it went off on a 45 odyssey trying to analyze too much. But other than those, it\u2019s been a pretty good daily workhorse. And you can\u2019t beat the price. Even Deepseek 4.1 Flash is mor", "context": "", "use_case": "daily coding model", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802082048, "item_id": 50130128, "url": "https://news.ycombinator.com/item?id=50130128", "by": "clx75", "date": "2026-10-17", "stance": "recommend", "quote": "I think now I can understand the person who created that thread.", "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": "implementing plans as agent", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802077776, "item_id": 50129861, "url": "https://news.ycombinator.com/item?id=50129861", "by": "gregwebs", "date": "2026-10-17", "stance": "recommend", "quote": "I am going back to DeepSeek for writing code.", "excerpt": "The company's own news page: https://api-docs.deepseek.com/news/news261017 They are showing this new version scoring significantly higher across multiple coding benchmarks, beating Sol 5.6 (6.1 is the latest) across 10/11 benchmarks. I am going back to DeepSeek for writing code. I was trying out using Sol and liking the result, but this should close a lot of that gap. I use Opus/Sol for an implementation plan and cod", "context": "", "use_case": "writing code", "switched_from": "Sol", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 802074016, "item_id": 50129626, "url": "https://news.ycombinator.com/item?id=50129626", "by": "nickludlam", "date": "2026-10-17", "stance": "recommend", "quote": "I find Deepseek in particular is really good with Svelte work.", "excerpt": "This and (the not quite yet open) Kimi K2.8 are my main coding engines, and I find them incredibly cost efficient. I find Deepseek in particular is really good with Svelte work. Right now I use each of them in their own respective coding harnesses, deepseek-harness and kimi-code, but I do wonder what they are like inside something like opencode.", "context": "", "use_case": "coding, Svelte work", "switched_from": "", "story": "DeepSeek-v4.1-Flash-1017"}, {"id": 801611681, "item_id": 50100730, "url": "https://news.ycombinator.com/item?id=50100730", "by": "wereHamster", "date": "2026-10-15", "stance": "recommend", "quote": "a minute later had my confirmation, along with a really nice explanation", "excerpt": "Last week I got a report about a vuln in one of our products. Asked Claude to confirm it. It refused, so I told it \"fuck you, I'm switching to a chinese model to help me out here\". It replied, I'm paraphrasing, that it's \"not offended, but doesn't like the tone of voice I used towards it, and still refuses to serve my request\". Fine, so be it. I sent ^D^D back to it, switched to deepseek-v4.1-flash from within openco", "context": "maintains a product receiving vulnerability reports", "use_case": "confirming vulnerability reports", "switched_from": "Claude", "story": "Wanted to download an ISO, ended up writing a kernel exploit"}, {"id": 801427744, "item_id": 50089234, "url": "https://news.ycombinator.com/item?id=50089234", "by": "ryan_glass", "date": "2026-10-14", "stance": "recommend", "quote": "I use Deepseek V4 flash 0731 Q8 quantisation from unsolth with reasoning turned off on a local server for quick easy tasks", "excerpt": "I thought it would be interesting to see if the local AI I use for quick question and answer stuff could crack the code. I use Deepseek V4 flash 0731 Q8 quantisation from unsolth with reasoning turned off on a local server for quick easy tasks. Note I use another model locally for harder tasks that need reasoning. I was a bit dissappointed that it knew it was Poe's cypher so didn't need to even try to crack the code.", "context": "", "use_case": "local quick Q&A tasks", "switched_from": "", "story": "The Gold-Bug: Edgar Allan Poe's story that taught me about cryptography (2024)"}, {"id": 800954288, "item_id": 50059643, "url": "https://news.ycombinator.com/item?id=50059643", "by": "joshheitzman", "date": "2026-10-12", "stance": "recommend", "quote": "Deepseek v4.1 Flash @ deepinfra.com has been really impressive and I've switched to that as my daily driver.", "excerpt": "Exactly. Source code is just a spec for a compiler / interpreter. My latest benchmark for FluxImp ( open sourced yesterday at https://fluximp.org ) purposely leaves in some ambiguity. Some model offerings (model at specific provider) consistently make better decisions to resolve the ambiguity than others and I want see that show up in the benchmark metrics as extra turns, extra cost, and possibly outright failure. De", "context": "", "use_case": "coding agent daily driver", "switched_from": "", "story": "The gap in shared understanding of LLM capability is widening"}, {"id": 800546449, "item_id": 50034153, "url": "https://news.ycombinator.com/item?id=50034153", "by": "AustinDev", "date": "2026-10-10", "stance": "recommend", "quote": "Running DeepSeek 4.1 Flash at FP4 is going to need at least a 10-15k machine. I'd consider that near frontier level.", "excerpt": ">You can already performantly run frontier level models on systems with $1600 tier video cards. Quanted out Qwen 3.8 models are not frontier level. Have you used them for real work? They suck. They are so unsure of themselves due to the quant-lobotomy. They just endlessly loop in thinking even when they have the right decision and then output it when they get to the thinking cap. They score close in benchmarks becaus", "context": "", "use_case": "", "switched_from": "", "story": "100+ reactions to 100+ solutions"}, {"id": 800403126, "item_id": 50025195, "url": "https://news.ycombinator.com/item?id=50025195", "by": "aftbit", "date": "2026-10-09", "stance": "warn", "quote": "they're more likely to go down rabbit holes, confidently assert falsehoods, or fix bugs with changes to my test harness rather than my code.", "excerpt": "I have had good luck rotating between the three tiers of GPT 5.6 with occasional jumps up to Astra or Fable. Most of my work is with GPT-5.6-Sol. Simple tasks like data extraction or trivial refactors (rename this variable etc) I often push down to Terra or Luna, or even to self-hosted Gemma4:31B. Very tricky stuff, like planning a new feature, design review, or code review of a complex change across multiple reposit", "context": "", "use_case": "easy coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800370880, "item_id": 50023180, "url": "https://news.ycombinator.com/item?id=50023180", "by": "zackangelo", "date": "2026-10-09", "stance": "recommend", "quote": "I've been loving DS4.1 Flash, it's been one of my daily drivers", "excerpt": "I've been loving DS4.1 Flash, it's been one of my daily drivers since we started testing it internally. We just launched it on our platform today (Mixlayer, https://mixlayer.com), promo code LAUNCH-DSV41F gets you some free credits if anyone wants to check it out.", "context": "", "use_case": "daily LLM use", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800358800, "item_id": 50022425, "url": "https://news.ycombinator.com/item?id=50022425", "by": "erans", "date": "2026-10-09", "stance": "recommend", "quote": "I've been using DS 4.1 Flash for a while now as my main workhorse. It's amazing.", "excerpt": "I've been using DS 4.1 Flash for a while now as my main workhorse. It's amazing. It's fast and it is also a GREAT debugger - it might not always find the best solution but it will definitely find the problem quickly. I'm also the CEO of a new company - LunaRoute - so if you want private, fixed cost, we run our own server in the US type of DeepSeek 4.1 Flash check out https://www.lunaroute.com If you want a trial, hit", "context": "CEO of LunaRoute", "use_case": "main workhorse and debugging", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800358288, "item_id": 50022393, "url": "https://news.ycombinator.com/item?id=50022393", "by": "robertheadley", "date": "2026-10-09", "stance": "recommend", "quote": "DeepSeek Flash 4.1 is generally what I use as as my workhorse.", "excerpt": "DeepSeek Flash 4.1 is generally what I use as as my workhorse. I use ChatGPT web to create the outline, then have Deep Seek built it out. Works generally pretty great.", "context": "", "use_case": "building out outlines into code", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800347056, "item_id": 50021691, "url": "https://news.ycombinator.com/item?id=50021691", "by": "epolanski", "date": "2026-10-09", "stance": "recommend", "quote": "I use DS 4.1 flash it all the time, i struggle to spend more than 1 euro per day on it", "excerpt": "> It's really not cheaper than frontier subscriptions. I use DS 4.1 flash it all the time, i struggle to spend more than 1 euro per day on it, even working all day long.", "context": "", "use_case": "daily coding work", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800315056, "item_id": 50019691, "url": "https://news.ycombinator.com/item?id=50019691", "by": "Primer81", "date": "2026-10-09", "stance": "recommend", "quote": "Its a daily driver for sure. not even worth considering anything else", "excerpt": "I've been waiting for something like this article for a month. The other AI companies are being left in the dust. I rip about 200 million tokens at least a day and spend maybe $3 with deepseek flash v4.1 for self hosted related coding tasks for ~20 projects i work on / maintain in parallel. Its a daily driver for sure. not even worth considering anything else, but maybe a locally run model on my GPU at the moment.", "context": "maintains ~20 projects", "use_case": "self-hosted related coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800250993, "item_id": 50015687, "url": "https://news.ycombinator.com/item?id=50015687", "by": "yuhmahp", "date": "2026-10-09", "stance": "recommend", "quote": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "excerpt": "We use Claude Fable to plan, and DeepSeek 4.1 Flash (hosted on DeepInfra) for everything else. Very cost effective.", "context": "", "use_case": "general coding tasks", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800234272, "item_id": 50014642, "url": "https://news.ycombinator.com/item?id=50014642", "by": "ltbarcly3", "date": "2026-10-09", "stance": "warn", "quote": "DeepSeek 4.1 Flash kindof sucks.", "excerpt": "DeepSeek 4.1 Flash kindof sucks. I used it a bunch and it kindof sucks. I don't know if they are gaming benchmarks or what. Luna is on par in benchmarks and my personal experience is Luna is better for what I do, and Luna is cheaper. Comparing Deepseek 4.1 flash to Opus is just ludicrous. https://artificialanalysis.ai/models/releases/comparisons?compare=gpt-6-luna%2Cdeepseek-v4-1-flash", "context": "", "use_case": "", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800204064, "item_id": 50012754, "url": "https://news.ycombinator.com/item?id=50012754", "by": "gregwebs", "date": "2026-10-08", "stance": "recommend", "quote": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast.", "excerpt": "I have been using DeepSeek 4.1 flash intensively for over a month. If I run it all day long it costs $1-2. Its fast. Previously I was always quickly running up to my Claude/Codex 5 hour window (on the $20/month plan). The cost savings of DeepSeek is real as shown in this article and I am using subsidized plans. DeepSeek is horrible at grilling sessions (the /grill* skills to make technical decisions). It doesn't know", "context": "", "use_case": "orchestrator, tester, subagent", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800198496, "item_id": 50012406, "url": "https://news.ycombinator.com/item?id=50012406", "by": "rapind", "date": "2026-10-08", "stance": "recommend", "quote": "I do think it's the most viable alternative I've seen so far", "excerpt": "It's really not cheaper than frontier subscriptions. It's getting closer, and it's a great model, but it is not more value per task than the frontier subscriptions. Don't be swayed by the token costs, it's very chatty, like 3x more tokens for the same task as sol. I used dsf 4.1 full time for about a week. It blows frontier API pricing out of the water, but again, look at cost per task, not token usage. Still easily ", "context": "", "use_case": "full-time coding work", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 800197696, "item_id": 50012356, "url": "https://news.ycombinator.com/item?id=50012356", "by": "apitman", "date": "2026-10-08", "stance": "recommend", "quote": "despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks", "excerpt": "> With my OpenCode Go sub of $10/month, DeepSeek is basically unlimited My OpenCode Go monthly window was scheduled to reset this morning. It was sitting at 22% used despite me using DeepSeek V4.1 Flash heavily as my implementation agent the past couple weeks (I use gpt-6.1-sol high for planning/orchestration). I had 1.5 hours left so I fired up first 10, then 20, and finally 50 concurrent subagents all working on re", "context": "", "use_case": "implementation agent, subagents", "switched_from": "", "story": "Why isn't the industry freaking out about DeepSeek 4.1 Flash?"}, {"id": 798302640, "item_id": 49893915, "url": "https://news.ycombinator.com/item?id=49893915", "by": "dools", "date": "2026-09-29", "stance": "recommend", "quote": "Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do.", "excerpt": "I go for weeks without topping up my Deepseek balance, and have to top up my OpenAI balance at least every week. GPT 6 Luna closed the gap significantly for sure (it seems to be about twice as expensive as DS v4.1 Flash), but Deepseek v4.1 Flash is still the best value model and capable enough for almost everything I need to do. Sometimes if it's babbling or can't nail down a solution I switch to Sol for one prompt, ", "context": "", "use_case": "planning and implementation", "switched_from": "", "story": "Sonnet 5.5"}, {"id": 795660832, "item_id": 49728802, "url": "https://news.ycombinator.com/item?id=49728802", "by": "surgical_fire", "date": "2026-09-16", "stance": "recommend", "quote": "I love DS flash, it is an amazing workhorse to implement plans created by more robust models", "excerpt": "What I find surprising is that DS Flash can do it at all. I love DS flash, it is an amazing workhorse to implement plans created by more robust models (such as GLM). But a more fair comparison would be of DS Flash with GLM Flash.", "context": "", "use_case": "implementing plans", "switched_from": "", "story": "DeepSeek v4.1 Flash Is Now Our Best Hacking Model"}]}