Free APILLMAPI keyOfficial
Cerebras Inference
Wafer-scale chips serving open models at extreme tokens-per-second, free tier
Cerebras Inference runs open-weight models (gpt-oss-120b, GLM, Llama, Qwen) on its wafer-scale hardware, delivering some of the highest throughput available. A free tier is available with no credit card via cloud.cerebras.ai, exposing an OpenAI-compatible chat endpoint.
Free tier
Free tier, no credit card required. Capped per-day token allowance; paid tiers raise limits.Rate limits
Free tier (verified June 2026): 5 RPM, 30,000 tokens/min, 1,000,000 tokens/day per model. Free model menu has included gpt-oss-120b and zai-glm-4.7. A reduced context cap may apply to some free models. Check the Limits page in your account for current values.Auth
API keyBase URL
https://api.cerebras.ai/v1What you'd build with it
- Latency-critical demos where raw generation speed is the wow factor
- Bursty automation steps that fit inside the 1M-tokens/day budget
- Benchmarking open models head-to-head on throughput
- Reasoning-heavy single-shot tasks where fast long outputs help
Key endpoints
- POST
/chat/completionsOpenAI-compatible chat completion (streaming supported) - POST
/completionsText completion - GET
/modelsList models available to your account
Members
Members get the runnable recipe
Sign in free to copy the example request, see a sample response, and read the gotchas.
Tags
llminferenceopenai-compatiblehigh-throughputopen-weights