{"slug": "cerebras", "name": "Cerebras", "category": "ai", "category_name": "AI models and assistants", "description": "AI inference provider running open-weight models on custom wafer-scale chips at very high speed", "recommenders": 9, "commenters": 13, "warners": 0, "score": 8.672, "rising": -0.29, "first": "2025-11-06", "latest": "2026-10-25", "why": "Loved for near-instant inference speed, which helps in many-turn agentic coding with hosted models like GLM 4.6.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 9, "neutral": 3, "compare": 1, "warn": 0}, "spark": [4, 2, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1], "listed": true, "loved": ["Speed", "Coding assistance"], "caveat": "Rate limits", "short_quote": "No affiliation with Cerebras, just a happy customer.", "icon": true, "domain": "cerebras.ai", "github_repo": null, "official_url": "https://www.cerebras.ai", "pricing": "freemium", "aliases": ["Cerebras"], "verdict": [{"text": "Loved for near-instant inference speed", "cites": [{"id": 803726480, "item_id": 50232905, "url": "https://news.ycombinator.com/item?id=50232905", "by": "nexus2045", "date": "2026-10-25", "stance": "recommend", "quote": "the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point", "excerpt": "This - I use Cerebras supported models here and there, the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point. Cerebras if you're reading this, get your act together - maybe that will also help your stock price. Give consumers more options, not just to OpenAI.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras Deploys Trillion-Parameter AI Models"}, {"id": 741769057, "item_id": 46360566, "url": "https://news.ycombinator.com/item?id=46360566", "by": "anonzzzies", "date": "2025-12-22", "stance": "recommend", "quote": "But seeing this generate at that speed is just jaw dropping.", "excerpt": "I have been using 4.6 on Cerebras (or Groq with other models) since it dropped and it is a glimpse of the future. If AGI never happens but we manage to optimise things so I can run that on my handheld/tablet/laptop device, I am beyond happy. And I guess that might happen. Maybe with custom inference hardware like Cerebras. But seeing this generate at that speed is just jaw dropping.", "context": "", "use_case": "fast model inference", "switched_from": "", "story": "GLM-4.7: Advancing the Coding Capability"}, {"id": 789746800, "item_id": 49359175, "url": "https://news.ycombinator.com/item?id=49359175", "by": "petesergeant", "date": "2026-08-19", "stance": "recommend", "quote": "It\u2019s every bit as fast and reliable for my needs as claimed.", "excerpt": "I use Cerebras via OpenRouter. It\u2019s every bit as fast and reliable for my needs as claimed. I suspect the reason is that they can either be making peanuts selling inference to plebs like me via OpenRouter, or making bank selling the more expensive models to businesses directly. In short: I would be very surprised if they have die capacity, and are at this point maximising revenue per chip.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras CS4"}, {"id": 746906722, "item_id": 46681670, "url": "https://news.ycombinator.com/item?id=46681670", "by": "esafak", "date": "2026-01-19", "stance": "recommend", "quote": "When I want fast I reach for Gemini, or Cerebras", "excerpt": "When I want fast I reach for Gemini, or Cerebras: https://www.cerebras.ai/blog/glm-4-7 GLM 4.7 is good enough to be a daily driver but it does frustrate me at times with poor instruction following.", "context": "", "use_case": "fast inference", "switched_from": "", "story": "GLM-4.7-Flash"}]}, {"text": ", which helps in many-turn agentic coding with hosted models like GLM 4.6.", "cites": [{"id": 733461745, "item_id": 45841359, "url": "https://news.ycombinator.com/item?id=45841359", "by": "ericd", "date": "2025-11-06", "stance": "recommend", "quote": "Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations", "excerpt": "Absolutely, especially the part about just rolling your own alternative to Claude Code - build your own lightsaber. Having your coding agent improve itself is a pretty magical experience. And then you can trivially swap in whatever model you want (Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations with big lumps of context, though gpt-oss 120b is obviously no", "context": "", "use_case": "fast inference for agents", "switched_from": "", "story": "You should write an agent"}, {"id": 733668481, "item_id": 45854280, "url": "https://news.ycombinator.com/item?id=45854280", "by": "divmain", "date": "2025-11-07", "stance": "recommend", "quote": "No affiliation with Cerebras, just a happy customer.", "excerpt": "I have been an AI-coding skeptic for some time. I always acknowledged LLMs as useful for solving specific problems and making certain things possible that weren't possible before. But I've not been surprised to see AI fail to live up to the hype. And I never had a personally magical moment - an experience that shifted my perspective \u00e0 la the peak end rule. I've been using GLM 4.6 on Cerebras for the last week or so, ", "context": "", "use_case": "fast inference", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 733714085, "item_id": 45857130, "url": "https://news.ycombinator.com/item?id=45857130", "by": "sheepscreek", "date": "2025-11-08", "stance": "recommend", "quote": "Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.", "excerpt": "I used their $50 plan and with the previously offered Qwen3 coder 480B. While fast - none of the \u201csupported\u201d tools I tried were able to use it in a way that didn\u2019t hit the per minute request limit in a few seconds. It was incredibly frustrating. For the record, I tried OpenCoder, VSCode, Quen Coder CLI, octofriend and a few others I don\u2019t remember. Fast forward to now, when GLM 4.6 has replaced Qwen3 coder in their s", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}]}, {"text": " Common complaints are rate limits and an outdated, frequently retired model lineup,", "cites": [{"id": 733714085, "item_id": 45857130, "url": "https://news.ycombinator.com/item?id=45857130", "by": "sheepscreek", "date": "2025-11-08", "stance": "recommend", "quote": "Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.", "excerpt": "I used their $50 plan and with the previously offered Qwen3 coder 480B. While fast - none of the \u201csupported\u201d tools I tried were able to use it in a way that didn\u2019t hit the per minute request limit in a few seconds. It was incredibly frustrating. For the record, I tried OpenCoder, VSCode, Quen Coder CLI, octofriend and a few others I don\u2019t remember. Fast forward to now, when GLM 4.6 has replaced Qwen3 coder in their s", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 803726480, "item_id": 50232905, "url": "https://news.ycombinator.com/item?id=50232905", "by": "nexus2045", "date": "2026-10-25", "stance": "recommend", "quote": "the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point", "excerpt": "This - I use Cerebras supported models here and there, the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point. Cerebras if you're reading this, get your act together - maybe that will also help your stock price. Give consumers more options, not just to OpenAI.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras Deploys Trillion-Parameter AI Models"}, {"id": 734943344, "item_id": 45933959, "url": "https://news.ycombinator.com/item?id=45933959", "by": "jpau", "date": "2025-11-14", "stance": "recommend", "quote": "I love Cerebras. I also love that they've started to scale rate limits to useful levels", "excerpt": "I love Cerebras. I also love that they've started to scale rate limits to useful levels (which is relatively new). I still don't know how long they'll support our chosen model. On Oct 22 I got an email saying that ``` - qwen-3-coder-480b will be available until Nov 5, 2025 - qwen-3-235b-a22b-thinking-2507 will be available until Nov 14, 2025 ``` That's not a lot of notice! I don't want to spend all my time benchmarki", "context": "builds product on hosted models", "use_case": "LLM inference", "switched_from": "", "story": "Why isn't everyone using Cerebras?"}]}, {"text": " plus high cost for agentic work without prompt caching or cache discounts.", "cites": [{"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 792873456, "item_id": 49554591, "url": "https://news.ycombinator.com/item?id=49554591", "by": "gardnr", "date": "2026-09-03", "stance": "neutral", "quote": "I used their Coding Plan for a few months.", "excerpt": "I used their Coding Plan for a few months. It is genuinely difficult to keep up with the models. The output is so fast. Qwen 3.8 27B is likely one of the strongest models they've hosted so far. Edit: it looks like this is only available on a API token pricing. Does anyone know if they have rolled out prompt caching yet? It used to get pretty expensive for agentic coding tasks with no prompt caching.", "context": "", "use_case": "agentic coding inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}]}], "praise": [{"label": "Speed", "summary": "Commenters repeatedly describe the inference speed as extremely fast or near instant.", "quotes": [{"id": 741769057, "item_id": 46360566, "url": "https://news.ycombinator.com/item?id=46360566", "by": "anonzzzies", "date": "2025-12-22", "stance": "recommend", "quote": "But seeing this generate at that speed is just jaw dropping.", "excerpt": "I have been using 4.6 on Cerebras (or Groq with other models) since it dropped and it is a glimpse of the future. If AGI never happens but we manage to optimise things so I can run that on my handheld/tablet/laptop device, I am beyond happy. And I guess that might happen. Maybe with custom inference hardware like Cerebras. But seeing this generate at that speed is just jaw dropping.", "context": "", "use_case": "fast model inference", "switched_from": "", "story": "GLM-4.7: Advancing the Coding Capability"}, {"id": 733461745, "item_id": 45841359, "url": "https://news.ycombinator.com/item?id=45841359", "by": "ericd", "date": "2025-11-06", "stance": "recommend", "quote": "Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations", "excerpt": "Absolutely, especially the part about just rolling your own alternative to Claude Code - build your own lightsaber. Having your coding agent improve itself is a pretty magical experience. And then you can trivially swap in whatever model you want (Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations with big lumps of context, though gpt-oss 120b is obviously no", "context": "", "use_case": "fast inference for agents", "switched_from": "", "story": "You should write an agent"}, {"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}]}, {"label": "Coding assistance", "summary": "Users report productive coding sessions with GLM models hosted on Cerebras.", "quotes": [{"id": 733668481, "item_id": 45854280, "url": "https://news.ycombinator.com/item?id=45854280", "by": "divmain", "date": "2025-11-07", "stance": "recommend", "quote": "No affiliation with Cerebras, just a happy customer.", "excerpt": "I have been an AI-coding skeptic for some time. I always acknowledged LLMs as useful for solving specific problems and making certain things possible that weren't possible before. But I've not been surprised to see AI fail to live up to the hype. And I never had a personally magical moment - an experience that shifted my perspective \u00e0 la the peak end rule. I've been using GLM 4.6 on Cerebras for the last week or so, ", "context": "", "use_case": "fast inference", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 733714085, "item_id": 45857130, "url": "https://news.ycombinator.com/item?id=45857130", "by": "sheepscreek", "date": "2025-11-08", "stance": "recommend", "quote": "Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.", "excerpt": "I used their $50 plan and with the previously offered Qwen3 coder 480B. While fast - none of the \u201csupported\u201d tools I tried were able to use it in a way that didn\u2019t hit the per minute request limit in a few seconds. It was incredibly frustrating. For the record, I tried OpenCoder, VSCode, Quen Coder CLI, octofriend and a few others I don\u2019t remember. Fast forward to now, when GLM 4.6 has replaced Qwen3 coder in their s", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 778084353, "item_id": 48630272, "url": "https://news.ycombinator.com/item?id=48630272", "by": "elzbardico", "date": "2026-06-22", "stance": "neutral", "quote": "I used it with Cerebras inference at a time when it had a good coding plan at a low price", "excerpt": "If you are a real engineer and uses the LLM as a pair programmer instead of delegating everything to it, even GLM 4.7 was already good enough to help you with a lot of work. I used it with Cerebras inference at a time when it had a good coding plan at a low price, and delivered tons of stuff using it.", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "GLM 5.2 vs. Opus"}]}, {"label": "Reliability", "summary": "One user says it is as fast and reliable as claimed when used via OpenRouter.", "quotes": [{"id": 789746800, "item_id": 49359175, "url": "https://news.ycombinator.com/item?id=49359175", "by": "petesergeant", "date": "2026-08-19", "stance": "recommend", "quote": "It\u2019s every bit as fast and reliable for my needs as claimed.", "excerpt": "I use Cerebras via OpenRouter. It\u2019s every bit as fast and reliable for my needs as claimed. I suspect the reason is that they can either be making peanuts selling inference to plebs like me via OpenRouter, or making bank selling the more expensive models to businesses directly. In short: I would be very surprised if they have die capacity, and are at this point maximising revenue per chip.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras CS4"}]}, {"label": "Improved rate limits", "summary": "A builder notes rate limits have recently been raised to useful levels.", "quotes": [{"id": 734943344, "item_id": 45933959, "url": "https://news.ycombinator.com/item?id=45933959", "by": "jpau", "date": "2025-11-14", "stance": "recommend", "quote": "I love Cerebras. I also love that they've started to scale rate limits to useful levels", "excerpt": "I love Cerebras. I also love that they've started to scale rate limits to useful levels (which is relatively new). I still don't know how long they'll support our chosen model. On Oct 22 I got an email saying that ``` - qwen-3-coder-480b will be available until Nov 5, 2025 - qwen-3-235b-a22b-thinking-2507 will be available until Nov 14, 2025 ``` That's not a lot of notice! I don't want to spend all my time benchmarki", "context": "builds product on hosted models", "use_case": "LLM inference", "switched_from": "", "story": "Why isn't everyone using Cerebras?"}]}], "caveats": [{"label": "Rate limits", "summary": "Users hit per-minute request limits or 429 errors during coding sessions.", "quotes": [{"id": 733714085, "item_id": 45857130, "url": "https://news.ycombinator.com/item?id=45857130", "by": "sheepscreek", "date": "2025-11-08", "stance": "recommend", "quote": "Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.", "excerpt": "I used their $50 plan and with the previously offered Qwen3 coder 480B. While fast - none of the \u201csupported\u201d tools I tried were able to use it in a way that didn\u2019t hit the per minute request limit in a few seconds. It was incredibly frustrating. For the record, I tried OpenCoder, VSCode, Quen Coder CLI, octofriend and a few others I don\u2019t remember. Fast forward to now, when GLM 4.6 has replaced Qwen3 coder in their s", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}]}, {"label": "Model deprecation", "summary": "Hosted models have been retired on short notice, disrupting products built on them.", "quotes": [{"id": 734943344, "item_id": 45933959, "url": "https://news.ycombinator.com/item?id=45933959", "by": "jpau", "date": "2025-11-14", "stance": "recommend", "quote": "I love Cerebras. I also love that they've started to scale rate limits to useful levels", "excerpt": "I love Cerebras. I also love that they've started to scale rate limits to useful levels (which is relatively new). I still don't know how long they'll support our chosen model. On Oct 22 I got an email saying that ``` - qwen-3-coder-480b will be available until Nov 5, 2025 - qwen-3-235b-a22b-thinking-2507 will be available until Nov 14, 2025 ``` That's not a lot of notice! I don't want to spend all my time benchmarki", "context": "builds product on hosted models", "use_case": "LLM inference", "switched_from": "", "story": "Why isn't everyone using Cerebras?"}]}, {"label": "Model selection", "summary": "Available models are called outdated or weaker than frontier models.", "quotes": [{"id": 803726480, "item_id": 50232905, "url": "https://news.ycombinator.com/item?id=50232905", "by": "nexus2045", "date": "2026-10-25", "stance": "recommend", "quote": "the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point", "excerpt": "This - I use Cerebras supported models here and there, the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point. Cerebras if you're reading this, get your act together - maybe that will also help your stock price. Give consumers more options, not just to OpenAI.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras Deploys Trillion-Parameter AI Models"}, {"id": 733461745, "item_id": 45841359, "url": "https://news.ycombinator.com/item?id=45841359", "by": "ericd", "date": "2025-11-06", "stance": "recommend", "quote": "Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations", "excerpt": "Absolutely, especially the part about just rolling your own alternative to Claude Code - build your own lightsaber. Having your coding agent improve itself is a pretty magical experience. And then you can trivially swap in whatever model you want (Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations with big lumps of context, though gpt-oss 120b is obviously no", "context": "", "use_case": "fast inference for agents", "switched_from": "", "story": "You should write an agent"}]}, {"label": "Cost", "summary": "Per-token pricing without cache discounts makes agentic coding expensive.", "quotes": [{"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 792873456, "item_id": 49554591, "url": "https://news.ycombinator.com/item?id=49554591", "by": "gardnr", "date": "2026-09-03", "stance": "neutral", "quote": "I used their Coding Plan for a few months.", "excerpt": "I used their Coding Plan for a few months. It is genuinely difficult to keep up with the models. The output is so fast. Qwen 3.8 27B is likely one of the strongest models they've hosted so far. Edit: it looks like this is only available on a API token pricing. Does anyone know if they have rolled out prompt caching yet? It used to get pretty expensive for agentic coding tasks with no prompt caching.", "context": "", "use_case": "agentic coding inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}]}, {"label": "Memory constraints", "summary": "One commenter flags a memory constraint that may limit suitability for some projects.", "quotes": [{"id": 743152560, "item_id": 46447035, "url": "https://news.ycombinator.com/item?id=46447035", "by": "jhoke", "date": "2025-12-31", "stance": "recommend", "quote": "take a look at Cerebras", "excerpt": "Hey, for less latency for the llm portion, take a look at Cerebras. It won't run OpenAI models, but if you can substitute for an equivalent you might be able to have better speeds. There is a memory constraint so not sure if it's the most suitable for the project though. Curious to see how this works out in terms of consistent durable performance.", "context": "", "use_case": "low-latency LLM inference", "switched_from": "", "story": "Ask HN: My multi-agent financial sentiment architecture"}]}], "who": [{"by": "jpau", "context": "builds product on hosted models", "url": "https://news.ycombinator.com/item?id=45933959", "date": "2025-11-14"}], "tags": ["for agentic coding", "for low-latency apps", "available via OpenRouter", "for small one-shot tasks"], "use_cases": ["fast llm inference", "fast inference for coding", "fast inference", "title generation", "agentic coding inference", "low-latency llm inference"], "compared": [{"slug": "glm", "name": "GLM", "co": 5, "switched_from": 0, "switched_to": 0}, {"slug": "openrouter", "name": "OpenRouter", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "claude-code", "name": "Claude Code", "co": 2, "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=49554520", "title": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=45852751", "title": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec", "mentions": 2}, {"url": "https://news.ycombinator.com/item?id=50227356", "title": "Cerebras Deploys Trillion-Parameter AI Models", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=49354949", "title": "Cerebras CS4", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=48626866", "title": "GLM 5.2 vs. Opus", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=46679872", "title": "GLM-4.7-Flash", "mentions": 1}], "sentiment_by_month": {"2025-11": {"recommend": 4}, "2025-12": {"recommend": 2}, "2026-01": {"recommend": 1}, "2026-02": {}, "2026-03": {}, "2026-04": {}, "2026-05": {}, "2026-06": {"neutral": 1}, "2026-07": {}, "2026-08": {"recommend": 1}, "2026-09": {"neutral": 2, "compare": 1}, "2026-10": {"recommend": 1}}, "quotes": [{"id": 803726480, "item_id": 50232905, "url": "https://news.ycombinator.com/item?id=50232905", "by": "nexus2045", "date": "2026-10-25", "stance": "recommend", "quote": "the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point", "excerpt": "This - I use Cerebras supported models here and there, the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point. Cerebras if you're reading this, get your act together - maybe that will also help your stock price. Give consumers more options, not just to OpenAI.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras Deploys Trillion-Parameter AI Models"}, {"id": 792952336, "item_id": 49559521, "url": "https://news.ycombinator.com/item?id=49559521", "by": "ma2kx", "date": "2026-09-04", "stance": "neutral", "quote": "small one shot task like title generation. At least thats why I use the free tier for.", "excerpt": "I guess Cerebras didnt intend the model for agentic coding but rather for small one shot task like title generation. At least thats why I use the free tier for.", "context": "", "use_case": "title generation", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 792891857, "item_id": 49555741, "url": "https://news.ycombinator.com/item?id=49555741", "by": "eli", "date": "2026-09-03", "stance": "compare", "quote": "cerebras was 5.6x more expensive in exchange for being 2.8x faster", "excerpt": "I just did a little anecdotal test. Had pi + cerebras review a recent commit and asked a few quick followups on it. Worked great. The Cerebras session cost me $1.60 and took a total of 5.1 mins. I did get a few brief 429 rate limit errors in there. The p50 speed was 890 tok/s and 0.64s TTFT. Using OpenRouter averages, that would've cost $0.29 (no cache discount at Cerebras!) and would've taken about 14.4 minutes. So ", "context": "former Cerebras coding plan user", "use_case": "fast LLM inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 792873456, "item_id": 49554591, "url": "https://news.ycombinator.com/item?id=49554591", "by": "gardnr", "date": "2026-09-03", "stance": "neutral", "quote": "I used their Coding Plan for a few months.", "excerpt": "I used their Coding Plan for a few months. It is genuinely difficult to keep up with the models. The output is so fast. Qwen 3.8 27B is likely one of the strongest models they've hosted so far. Edit: it looks like this is only available on a API token pricing. Does anyone know if they have rolled out prompt caching yet? It used to get pretty expensive for agentic coding tasks with no prompt caching.", "context": "", "use_case": "agentic coding inference", "switched_from": "", "story": "Qwen 3.8 27B available on Cerebras at 1500 tokens/s"}, {"id": 789746800, "item_id": 49359175, "url": "https://news.ycombinator.com/item?id=49359175", "by": "petesergeant", "date": "2026-08-19", "stance": "recommend", "quote": "It\u2019s every bit as fast and reliable for my needs as claimed.", "excerpt": "I use Cerebras via OpenRouter. It\u2019s every bit as fast and reliable for my needs as claimed. I suspect the reason is that they can either be making peanuts selling inference to plebs like me via OpenRouter, or making bank selling the more expensive models to businesses directly. In short: I would be very surprised if they have die capacity, and are at this point maximising revenue per chip.", "context": "", "use_case": "fast LLM inference", "switched_from": "", "story": "Cerebras CS4"}, {"id": 778084353, "item_id": 48630272, "url": "https://news.ycombinator.com/item?id=48630272", "by": "elzbardico", "date": "2026-06-22", "stance": "neutral", "quote": "I used it with Cerebras inference at a time when it had a good coding plan at a low price", "excerpt": "If you are a real engineer and uses the LLM as a pair programmer instead of delegating everything to it, even GLM 4.7 was already good enough to help you with a lot of work. I used it with Cerebras inference at a time when it had a good coding plan at a low price, and delivered tons of stuff using it.", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "GLM 5.2 vs. Opus"}, {"id": 746906722, "item_id": 46681670, "url": "https://news.ycombinator.com/item?id=46681670", "by": "esafak", "date": "2026-01-19", "stance": "recommend", "quote": "When I want fast I reach for Gemini, or Cerebras", "excerpt": "When I want fast I reach for Gemini, or Cerebras: https://www.cerebras.ai/blog/glm-4-7 GLM 4.7 is good enough to be a daily driver but it does frustrate me at times with poor instruction following.", "context": "", "use_case": "fast inference", "switched_from": "", "story": "GLM-4.7-Flash"}, {"id": 743152560, "item_id": 46447035, "url": "https://news.ycombinator.com/item?id=46447035", "by": "jhoke", "date": "2025-12-31", "stance": "recommend", "quote": "take a look at Cerebras", "excerpt": "Hey, for less latency for the llm portion, take a look at Cerebras. It won't run OpenAI models, but if you can substitute for an equivalent you might be able to have better speeds. There is a memory constraint so not sure if it's the most suitable for the project though. Curious to see how this works out in terms of consistent durable performance.", "context": "", "use_case": "low-latency LLM inference", "switched_from": "", "story": "Ask HN: My multi-agent financial sentiment architecture"}, {"id": 741769057, "item_id": 46360566, "url": "https://news.ycombinator.com/item?id=46360566", "by": "anonzzzies", "date": "2025-12-22", "stance": "recommend", "quote": "But seeing this generate at that speed is just jaw dropping.", "excerpt": "I have been using 4.6 on Cerebras (or Groq with other models) since it dropped and it is a glimpse of the future. If AGI never happens but we manage to optimise things so I can run that on my handheld/tablet/laptop device, I am beyond happy. And I guess that might happen. Maybe with custom inference hardware like Cerebras. But seeing this generate at that speed is just jaw dropping.", "context": "", "use_case": "fast model inference", "switched_from": "", "story": "GLM-4.7: Advancing the Coding Capability"}, {"id": 734943344, "item_id": 45933959, "url": "https://news.ycombinator.com/item?id=45933959", "by": "jpau", "date": "2025-11-14", "stance": "recommend", "quote": "I love Cerebras. I also love that they've started to scale rate limits to useful levels", "excerpt": "I love Cerebras. I also love that they've started to scale rate limits to useful levels (which is relatively new). I still don't know how long they'll support our chosen model. On Oct 22 I got an email saying that ``` - qwen-3-coder-480b will be available until Nov 5, 2025 - qwen-3-235b-a22b-thinking-2507 will be available until Nov 14, 2025 ``` That's not a lot of notice! I don't want to spend all my time benchmarki", "context": "builds product on hosted models", "use_case": "LLM inference", "switched_from": "", "story": "Why isn't everyone using Cerebras?"}, {"id": 733714085, "item_id": 45857130, "url": "https://news.ycombinator.com/item?id=45857130", "by": "sheepscreek", "date": "2025-11-08", "stance": "recommend", "quote": "Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.", "excerpt": "I used their $50 plan and with the previously offered Qwen3 coder 480B. While fast - none of the \u201csupported\u201d tools I tried were able to use it in a way that didn\u2019t hit the per minute request limit in a few seconds. It was incredibly frustrating. For the record, I tried OpenCoder, VSCode, Quen Coder CLI, octofriend and a few others I don\u2019t remember. Fast forward to now, when GLM 4.6 has replaced Qwen3 coder in their s", "context": "", "use_case": "fast inference for coding", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 733668481, "item_id": 45854280, "url": "https://news.ycombinator.com/item?id=45854280", "by": "divmain", "date": "2025-11-07", "stance": "recommend", "quote": "No affiliation with Cerebras, just a happy customer.", "excerpt": "I have been an AI-coding skeptic for some time. I always acknowledged LLMs as useful for solving specific problems and making certain things possible that weren't possible before. But I've not been surprised to see AI fail to live up to the hype. And I never had a personally magical moment - an experience that shifted my perspective \u00e0 la the peak end rule. I've been using GLM 4.6 on Cerebras for the last week or so, ", "context": "", "use_case": "fast inference", "switched_from": "", "story": "Cerebras Code now supports GLM 4.6 at 1000 tokens/sec"}, {"id": 733461745, "item_id": 45841359, "url": "https://news.ycombinator.com/item?id=45841359", "by": "ericd", "date": "2025-11-06", "stance": "recommend", "quote": "Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations", "excerpt": "Absolutely, especially the part about just rolling your own alternative to Claude Code - build your own lightsaber. Having your coding agent improve itself is a pretty magical experience. And then you can trivially swap in whatever model you want (Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations with big lumps of context, though gpt-oss 120b is obviously no", "context": "", "use_case": "fast inference for agents", "switched_from": "", "story": "You should write an agent"}]}