Skip to content
aicoolies logo

Cerebras Review: Wafer-Scale LLM Inference, API, and Pricing

A documented review of Cerebras Inference: the CS-3 wafer-scale hardware, the OpenAI-compatible API, the model catalog, and published pricing. Speed figures come from Cerebras and from Artificial Analysis, not from an aicoolies measurement.

reviewed by Raşit Akyol September 1, 2026 updated September 28, 2026

Documented evidence

rubric editorial-review-v1

This review is grounded in documented sources and repository analysis. It does not claim a unique hands-on reproducibility record.

Sources checked

Verdict

Cerebras is a strong fit when generation speed is the main constraint, such as voice agents, interactive coding assistants, and multi-step agent loops, and when the open-weights models it serves are good enough for the task. Artificial Analysis ranked Cerebras fastest of 18 providers for gpt-oss-120b (72-hour median, as of 28 Sep 2026); aicoolies has not measured its speed. Teams that need proprietary frontier models or very long context windows should compare alternatives first.

92/100

overall

Speed98
Privacy91
Dev Experience90

Verdict and evidence boundary

Cerebras sells LLM inference running on its own wafer-scale hardware. Cerebras describes the WSE-3 in its CS-3 system as a single wafer-scale chip with 46,225 mm² of silicon, 900,000 cores, and 44 GB of on-chip SRAM, and that design is the basis for its speed claims.

This review is Documented. It summarizes Cerebras documentation, its pricing page, its published performance figures, and the independent Artificial Analysis leaderboard. aicoolies has not run its own latency or throughput tests, so every performance figure below is attributed to its source.

Hardware Architecture: The Wafer-Scale Engine (CS-3)

On GPU clusters, token-by-token generation is often limited by moving model weights and KV cache between external HBM memory and compute. Cerebras keeps weights and activations in on-chip SRAM on one wafer-sized chip, which removes most off-chip memory traffic. Cerebras's datasheet lists 21 PB/s of on-chip memory bandwidth for the WSE-3, and the company claims inference up to 30 times faster than GPU-based systems.

Published Speed Figures

Cerebras lists about 3,000 tokens per second for gpt-oss-120b and about 1,850 tokens per second for Qwen 3.8 27B on its pricing page and in its documentation (checked 28 Sep 2026). Independently, Artificial Analysis ranked Cerebras fastest of 18 providers for gpt-oss-120b (72-hour median, as of 28 Sep 2026). These figures change as models and hardware are updated, so readers should check both sources for current numbers. aicoolies did not measure time-to-first-token, throughput, or end-to-end latency.

Developer Ergonomics and OpenAI SDK Compatibility

Cerebras documents an OpenAI-compatible API. Teams using the OpenAI SDK point their client at `https://api.cerebras.ai/v1` and supply a Cerebras API key (`client = OpenAI(base_url='https://api.cerebras.ai/v1', api_key=...)`).

The documentation covers tool calling, structured outputs with JSON schema, streaming, and sampling parameters such as temperature (0 to 2) and top_p (0 to 1). One documented limitation is that `json_object` mode cannot be combined with streaming. This makes Cerebras usable from LangChain, LlamaIndex, Pydantic AI, and custom agent loops through their OpenAI-compatible clients. aicoolies has not measured schema adherence or tool-call reliability.

Pricing, Model Catalog, and Trade-offs

The Developer tier is pay-as-you-go and starts with a free $5 credit. Cerebras lists gpt-oss-120b at $0.35 per million input tokens and $0.75 per million output tokens, and Qwen 3.8 27B at $0.99 and $1.49; Enterprise pricing is by quote (checked 28 Sep 2026).

The main trade-off is catalog scope. Cerebras's own API serves open-weights models rather than proprietary models such as Claude. Context limits vary by model and plan: gpt-oss-120b is listed at 65K tokens on the free tier and 131K on paid tiers, and Qwen 3.8 27B at 64K and 128K. Teams that need very long contexts should check these limits before committing.

Final Verdict

Cerebras is worth evaluating first when generation speed is the main constraint and an open-weights model fits the task. Artificial Analysis ranked it fastest of 18 providers for gpt-oss-120b (72-hour median, as of 28 Sep 2026), and the OpenAI-compatible API keeps switching cost low, so a team can time it on its own prompts before committing. Its aicoolies score is 92/100 (Speed 98, Privacy 91, Developer Experience 90). This is a Documented score based on published sources, not an aicoolies measurement.

Pros

  • High generation speed: Cerebras's pricing table lists about 3,000 tokens per second for gpt-oss-120b (checked 28 Sep 2026).
  • Full drop-in compatibility with the standard OpenAI Python and TypeScript SDKs via custom base_url.
  • Published per-token pricing, for example $0.35 per 1M input tokens and $0.75 per 1M output tokens for gpt-oss-120b (checked 28 Sep 2026).
  • Fast generation can shorten multi-step agent loops, where many sequential model calls add up.

Cons

  • The shared catalog serves open-weights models such as gpt-oss-120b, with no proprietary frontier models; Llama models are now offered only on Dedicated.
  • Context window sizes are currently constrained compared to hyperscaler 1M+ token GPU clusters.
  • Custom LoRA adapter serving and fine-tuned weight deployment are still expanding in cloud preview.

View Cerebras on aicoolies

Pricing, platforms, and community stacks — explore the full tool page

Alternatives to Cerebras

Ultra-fast LPU inference for open-weight models

Groq is an AI inference provider built around custom Language Processing Unit (LPU) hardware for low-latency open-weight model serving. GroqCloud exposes an OpenAI-compatible API for GPT-OSS, Qwen and Whisper models, with Llama and MiniMax models available through enterprise sales, plus model-specific rate limits, Batch and Flex processing, and usage-based pricing.

freemium

Open-weight inference, fine-tuning, and GPU-cloud platform

Together AI is a cloud platform for running, fine-tuning, batching, and training open-weight AI models. It supports serverless inference, dedicated endpoints, LoRA and full fine-tuning, GPU clusters, code-execution sandboxes, and async batch jobs up to 30B tokens per model. Current docs list fast-moving families such as Qwen, Kimi, GLM, GPT-OSS, DeepSeek, Llama, MiniMax, and Mistral.

freemium

Production-grade inference with serverless and on-demand GPUs

High-performance inference platform serving open-source and custom AI models at global scale, processing 13+ trillion tokens daily at ~180K requests per second. Fireworks AI delivers 1,000+ tokens per second on large models through quantization-aware tuning and adaptive speculation, with serverless, fine-tuning, and dedicated GPU options across text, image, and audio modalities.

freemium

FAQ

How does Cerebras achieve its high inference speeds?

Cerebras attributes its speed to the WSE-3 chip in its CS-3 system: a single wafer-scale chip with 900,000 AI cores and 44 GB of on-chip SRAM with 21 PB/s of memory bandwidth, which avoids the external memory bottlenecks of standard GPUs.

Can I use the OpenAI Python or Node.js SDK to connect to Cerebras?

Yes. Cerebras documents an OpenAI-compatible API. You set base_url to https://api.cerebras.ai/v1 and provide a Cerebras API key.

What are the pricing rates for Cerebras inference?

Cerebras publishes per-model rates. For gpt-oss-120b, its pricing table lists $0.35 per 1M input tokens and $0.75 per 1M output tokens (checked 28 Sep 2026). Llama models are now offered on Dedicated rather than in the shared catalog.

Which use cases benefit most from Cerebras ultra-fast inference?

Latency-sensitive workloads benefit most: real-time voice agents, interactive coding assistants, and multi-turn agent loops where many sequential model calls add up.

Sources & verification

Sources checked
Content verified

Verification dates are editorial checks. Routine CMS saves and automatic updatedAt timestamps do not advance them.