Our take
Cerebras is another custom-chip inference provider, using custom chips rather than standard graphics chips. It has an even smaller catalogue than Groq and is aimed at maximum single-stream speed on a handful of supported models.
Use this for maximum single-stream speed on one of its three supported models, or real-time applications where generation speed is the product feature. Skip it if you need a broad catalogue, competitive pricing, or verified EU data residency.
- Purpose-built wafer-scale inference hardware designed for very high tokens per second.
- Only three tracked offers, a very small catalogue.
- Premium pricing on its models compared with other hosts for comparable downloadable models.
Point your tools here
We have not recorded what a router needs for Cerebras yet — no base URL and no docs link on file. That is our gap, not a sign Cerebras has no API; its own documentation is the place to look until we close it.
Privacy & data handling
These answers cover requests sent to Cerebras through OpenRouter, as OpenRouter records them, last read 59 min ago. For Cerebras's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.
We have not dated this attestation or linked the report yet.
Prompt retention: none.