Cerebras
Ultra-fast AI inference on custom Wafer-Scale Engine chips. Up to 3000 tok/s output speed, 20x faster than GPU-based providers.
At a glance: Cerebras has 2 models listed on Inference Hub. The cheapest text model is GPT OSS 120B at $0.350 input and $0.750 output per 1M tokens, and the most expensive is Qwen3.8 27B at $0.990 input and $1.49 output per 1M tokens. Free tier: $5 credit that expires 30 days after activation; payment method required. Prices are checked against Cerebras's own pricing every week.
Features
Compliance
None listed
Pricing Model
per tokenDetails
https://api.cerebras.ai/v1 Model Catalog (4)
| Model | Type | Input $/1M | Output $/1M | Context | Speed | Status |
|---|---|---|---|---|---|---|
| GLM 4.7 Zhipu AI | llm | $2.25 | $2.75 | — | — | |
| GPT OSS 120B OpenAI · 117B MoE (5.1B active) | llm | $0.350 | $0.750 | — | — | |
| Qwen 3 235B Alibaba · 235B MoE | llm | $0.600 | $1.20 | — | — | |
| Qwen3.8 27B Alibaba | llm | $0.990 | $1.49 | 66k | — |
Cerebras guides
Zero Data Retention for AI APIs: Which Providers Store Your Prompts, and How to Turn It Off
Which AI API providers store prompts, for how long, and how to get zero data retention at OpenAI, Anthropic, Google, Azure, Bedrock, OpenRouter and 20 hosts.
Nov 2026AI APIs with Free Tiers in 2026: The Complete Guide
Every AI API with a free tier and its limits: Gemini, Groq, Cloudflare, Together AI, OpenRouter, fal.ai and more. Build with LLM and image APIs for $0.
Cerebras API FAQ
Is the Cerebras API free?
Yes, Cerebras has a free tier: $5 credit that expires 30 days after activation; payment method required. Paid usage is billed per model.
What is the cheapest Cerebras model?
GPT OSS 120B at $0.35 in / $0.75 out per 1M tokens, based on the 4 models we track for Cerebras.
How many models does Cerebras offer?
We track 4 Cerebras models: 4 LLM.
Is the Cerebras API OpenAI-compatible?
Yes. Cerebras exposes an OpenAI-compatible endpoint, so you can use the OpenAI SDKs by changing the base URL and API key.
Run Cerebras? Add a "Prices tracked on Inference Hub" badge to your site, or email hello@inferencehub.org if a price here is wrong.