# Cerebras Inference Hacker Approved page: https://hackerapproved.com/p/cerebras/ 63 people recommended Cerebras Inference in Hacker News comments, 33 had mixed feelings and 47 criticized it, November 2025 to October 2026. It was recommended in 12 of 12 months, in 55 threads. It does not meet the Hacker Approved badge rule. **Verdict:** People recommend it for extremely fast token generation that makes agents and coding feel near-instant, but many report harsh rate limits, capacity shortages and a limited, outdated model lineup. Summary written by Claude from 47 comments. It was written before the November 2026 recount, and 45 of those comments still count. Claude agents reviewed it against the quotes on 2 Nov 2026. ## Why people recommend it - Extremely fast token generation (12 comments in the sample): "the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point" ([source](https://news.ycombinator.com/item?id=50232905)) "With Cerebras it’s not even worth switching tabs, because it’ll finish the same task in like three seconds." ([source](https://news.ycombinator.com/item?id=45844741)) "I use them heavily in production because OSS 120B on Groq/Cerebras is one of the fastest models available" ([source](https://news.ycombinator.com/item?id=47212085)) - Speed transforms agentic coding workflows (3 comments in the sample): "use cerebras with claude code and watch it generate millions of tokens in a few minutes" ([source](https://news.ycombinator.com/item?id=47004315)) "Cerebras running Sol at 750 tokens per second, the harness "speed" becomes irrelevant" ([source](https://news.ycombinator.com/item?id=49292589)) "1500 tok/s is an impressive speed, and that speed is very important for agent tasks" ([source](https://news.ycombinator.com/item?id=49567196)) - Good value for heavy usage (3 comments in the sample): "Cheap enough for now, but of all the companies selling inference at a loss, Cerebras and Groq are probably losing the most per token" ([source](https://news.ycombinator.com/item?id=45855972)) "most fixed cloud subscriptions will run out really quickly and will be insufficient (Cerebras being an exception" ([source](https://news.ycombinator.com/item?id=46216192)) "24M tokens per day for 50 USD seems like good value even with 128k context limitation" ([source](https://news.ycombinator.com/item?id=45855806)) - Generous subscription token allowances (1 comment in the sample): "I love Cerebras. I also love that they've started to scale rate limits to useful levels" ([source](https://news.ycombinator.com/item?id=45933959)) ## What people warn about - Aggressive rate limits, unreliable service (4 comments in the sample): "they aggressively rate limit their APIs to the point where if you just run a single coding agent, your coding session will constantly stall out" ([source](https://news.ycombinator.com/item?id=50233819)) "they aren't reliable at all, because they have to share limited resources dedicated to sales" ([source](https://news.ycombinator.com/item?id=50235615)) "the $50/month or pay as you go was just not reliable…" ([source](https://news.ycombinator.com/item?id=47229121)) - Limited or outdated model selection (3 comments in the sample): "They do offer API services to individual users… though with a set of models that makes it unlikely that you want to use it." ([source](https://news.ycombinator.com/item?id=49355955)) "I like Cerabras, but I really wish they would make more of their hosted models generally available." ([source](https://news.ycombinator.com/item?id=48837080)) "Cerebras are only serving kimi for dedicated endpoint customers; for that you need a >$5m annual deal with them" ([source](https://news.ycombinator.com/item?id=48369285)) - Capacity constrained, enterprise gets priority (2 comments in the sample): "They're pretty supply constrained right now though and their production costs seem prohibitive." ([source](https://news.ycombinator.com/item?id=48665331)) "Cerebras is super constrained on capacity right now, all the support is going to enterprise customers." ([source](https://news.ycombinator.com/item?id=49559282)) - Context, output caps, no caching (2 comments in the sample): "Cerebras currently does not provide any discounts for prefix caching making its use for agentic workloads sqr(n_turns) more expensive." ([source](https://news.ycombinator.com/item?id=48447603)) "Their architecture recomputes every time so at 150k context every request will have to spend 1.5 min waiting" ([source](https://news.ycombinator.com/item?id=49560703)) ## Other LLM APIs people recommend - [Mistral](https://hackerapproved.com/p/mistral-large-4/): 206 recommend, 81 mixed, 205 criticize - [DeepSeek API](https://hackerapproved.com/p/deepseek-api/): 147 recommend, 17 mixed, 49 criticize - [OpenAI API](https://hackerapproved.com/p/openai-models/): 115 recommend, 41 mixed, 260 criticize - [Anthropic API](https://hackerapproved.com/p/anthropic-api/): 39 recommend, 18 mixed, 118 criticize ## Where to get it - [Visit the official site](https://www.cerebras.ai/) ## The badge A product earns the Hacker Approved badge when at least 10 people recommend it, recommenders outnumber critics at least 2 to 1 (mixed feelings count toward neither side), it was recommended in at least 4 of the 12 months and in at least 5 different threads, and no single thread supplies more than half of the people who recommend it. Health and money products are never badged. ## Cite as Hacker Approved, "Cerebras Inference", https://hackerapproved.com/p/cerebras/ (63 recommend, 33 mixed, 47 criticize; Hacker News comments, November 2025 to October 2026). Counts updated 2 Nov 2026, with data through October 31, 2026.