Cerebras
AI inference provider running open-weight models on custom wafer-scale chips at very high speed
Loved for near-instant inference speed1234, which helps in many-turn agentic coding with hosted models like GLM 4.6.567 Common complaints are rate limits and an outdated, frequently retired model lineup,891011 plus high cost for agentic work without prompt caching or cache discounts.1213
Why people recommend it
Speed · 3
Commenters repeatedly describe the inference speed as extremely fast or near instant.
“But seeing this generate at that speed is just jaw dropping.”
“Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations”
Show 1 more
“cerebras was 5.6x more expensive in exchange for being 2.8x faster”
Coding assistance · 3
Users report productive coding sessions with GLM models hosted on Cerebras.
“No affiliation with Cerebras, just a happy customer.”
“Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.”
Show 1 more
“I used it with Cerebras inference at a time when it had a good coding plan at a low price”
Reliability · 1
One user says it is as fast and reliable as claimed when used via OpenRouter.
“It’s every bit as fast and reliable for my needs as claimed.”
Improved rate limits · 1
A builder notes rate limits have recently been raised to useful levels.
“I love Cerebras. I also love that they've started to scale rate limits to useful levels”
Caveats and dissent
Rate limits · 2
Users hit per-minute request limits or 429 errors during coding sessions.
“Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.”
“cerebras was 5.6x more expensive in exchange for being 2.8x faster”
Model deprecation · 1
Hosted models have been retired on short notice, disrupting products built on them.
“I love Cerebras. I also love that they've started to scale rate limits to useful levels”
Model selection · 2
Available models are called outdated or weaker than frontier models.
“the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point”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 is crazy fast, for example, which makes a big difference for these many-turn tool call conversations”
Cost · 2
Per-token pricing without cache discounts makes agentic coding expensive.
“cerebras was 5.6x more expensive in exchange for being 2.8x faster”
“I used their Coding Plan for a few months.”
Memory constraints · 1
One commenter flags a memory constraint that may limit suitability for some projects.
“take a look at Cerebras”
Often compared with
Who recommends it in their own words
Threads where it came up
- Qwen 3.8 27B available on Cerebras at 1500 tokens/s 3 mentions
- Cerebras Code now supports GLM 4.6 at 1000 tokens/sec 2 mentions
- Cerebras Deploys Trillion-Parameter AI Models 1 mention
- Cerebras CS4 1 mention
- GLM 5.2 vs. Opus 1 mention
- GLM-4.7-Flash 1 mention
Data: cerebras.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 13
“the speed is amazing and I can see the potential with having almost instant inference, but the models are so outdated at this point”
“small one shot task like title generation. At least thats why I use the free tier for.”
“cerebras was 5.6x more expensive in exchange for being 2.8x faster”
“I used their Coding Plan for a few months.”
“It’s every bit as fast and reliable for my needs as claimed.”
“I used it with Cerebras inference at a time when it had a good coding plan at a low price”
“When I want fast I reach for Gemini, or Cerebras”
“take a look at Cerebras”
“But seeing this generate at that speed is just jaw dropping.”
“I love Cerebras. I also love that they've started to scale rate limits to useful levels”
“Cerebras + GLM 4.6 feels like Grok Fast 1 on steroids.”
“No affiliation with Cerebras, just a happy customer.”
“Cerebras is crazy fast, for example, which makes a big difference for these many-turn tool call conversations”