Cerebras API pricing
Updated
Cerebras lists standard prices for 2 models. The cheapest is OpenAI GPT OSS 120B at USD 0.35 input and USD 0.75 output per 1M tokens; the dearest is Qwen3.8 27B at USD 0.99 / USD 1.49. It publishes no cached-input discount (cached tokens bill at the input price), and lists no per-model batch, flex or priority prices on the pages we read.
- Models with a standard price
- 2
- Median input / output
- USD 0.67 / USD 1.12
- Cached-input prices
- 0 of 2
- Cheapest host for
- 0 of 2 shared models
Which models is Cerebras the cheapest provider for?
None. Every one of the 2 models that Cerebras shares with other providers is cheaper elsewhere at the standard tier. The closest is Qwen3.8 27B, where Cerebras ranks 10 of 15. It is the dearest provider for 1.
How does Cerebras price the most widely hosted models?
The 2 models Cerebras sells that the most other providers also sell, with Cerebras' rank among them on the 3:1 blend.
| Model | Cerebras | Rank | Cheapest |
|---|---|---|---|
| OpenAI GPT OSS 120B | USD 0.35 / USD 0.75 | 20 of 21 | CoreWeave Forge USD 0.03 / USD 0.17 |
| Qwen3.8 27B | USD 0.99 / USD 1.49 | 10 of 15 | Chutes USD 0.039 / USD 1.25 |
Which model makers does Cerebras sell?
Cerebras prices models from 2 makers. OpenAI leads with 1 model. Standard prices, USD per 1M input / output tokens.
| Maker | Models | Cheapest model | Median input / output |
|---|---|---|---|
| OpenAI | 1 | OpenAI GPT OSS 120B USD 0.35 / USD 0.75 | — |
| Qwen | 1 | Qwen3.8 27B USD 0.99 / USD 1.49 | — |
Cerebras price list
USD per 1M tokens: 2 prices for 2 models, one row per model with any other service tiers stacked under the standard price. Sources checked 2026-11-14 unless a row says otherwise. "vs maker" compares a standard price with the model maker's own API on the 3:1 blend; a minus means Cerebras is cheaper. Of 1 hosted model price with a maker price to compare, 1 is below the maker's and 0 above.
| Model | Tier | Input | Cached | Output | Context | vs maker | Source |
|---|---|---|---|---|---|---|---|
| OpenAI GPT OSS 120B | standard | USD 0.35 | USD 0.35 | USD 0.75 | 131,072 | — | source |
| Qwen3.8 27B | standard | USD 0.99 | USD 0.99 | USD 1.49 | 131,072 | −1% | source |
- Cached "—" means Cerebras' page shows no cache price for that row; budget cached tokens at the input price.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What service tiers does Cerebras publish?
Standard prices only: Cerebras publishes no per-model batch, flex, priority or fast price on the pages we read.
Cached input: Cerebras' pages state that cached tokens bill at the standard input rate, so there is no cache discount.
What we recorded about Cerebras' pricing page: Only two models have a public Developer-tier token price; Enterprise pricing is quote-only. Cerebras docs state cached tokens are billed at the standard input rate, so cached = input.
Is Cerebras cheaper than the model makers' own APIs?
Yes. Of 1 standard price for models whose maker sells its own API, 1 is below the maker's price, 0 above and 0 the same, on the 3:1 blend. The median difference is 1% below the maker. The biggest saving is Qwen3.8 27B, 1% below Alibaba Cloud.
BroadNet Inference and Cerebras
BroadRouter is BroadNet's own site, so BroadNet is never ranked on this page. BroadNet Inference also serves 1 model that Cerebras prices: Qwen3.8 27B at USD 0.175 / USD 2.17 against Cerebras' USD 0.99 / USD 1.49 (BroadNet is cheaper on the 3:1 blend). BroadNet prices are from its rate card, synced 2026-11-14. BroadNet Inference pricing.
More on Cerebras
- Cerebras' own pricing page (the source; checked 2026-11-14)
- LLM API free tiers and rate limits: Cerebras' free tier and default limits next to other providers'
- All 41 providers compared · Cheapest LLM APIs right now · Download every price (CSV, JSON)
Quick answers
Does Cerebras keep API prompts and outputs?
No. The Cerebras privacy policy says "we do not retain inputs and outputs associated with our training, inference, and chatbot Services" (Cerebras privacy policy, checked 2026-11-14).
Does Cerebras train on API data?
No. The Cerebras Terms of Service say using Service Content to provide the Service "does not grant Cerebras the right to use Service Content for the purpose of training or fine-tuning models" (Cerebras Terms of Service, checked 2026-11-14).
Does Cerebras have a free tier?
5 USD in Free Trial credits after adding a verified payment method; they expire 30 days after they are granted. Default limits (Free Trial, gpt-oss-120b): 5 requests a minute and 30,000 tokens a minute. Free Trial: 30K uncached TPM (90K total TPM) and 1M tokens a day. qwen-3.8-27b: 300 RPM (Cerebras rate limits docs, checked 2026-11-14).
Questions
How much does the Cerebras API cost per token?
Cerebras' standard prices run from USD 0.35 input and USD 0.75 output per 1M tokens (OpenAI GPT OSS 120B) to USD 0.99 / USD 1.49 (Qwen3.8 27B). The median across 2 standard prices is USD 0.67 input and USD 1.12 output per 1M, which is USD 0.00000067 per input token. Input, cached input and output are billed separately.
What is the cheapest model on Cerebras?
OpenAI GPT OSS 120B, at USD 0.35 input and USD 0.75 output per 1M tokens (USD 0.45 on a 3:1 input-to-output blend). The cheapest from other makers: Qwen3.8 27B (Qwen) at USD 0.99 / USD 1.49.
Is Cerebras cheaper than other providers?
It depends on the model. Of 2 models that other providers also price, Cerebras is the cheapest standard price for none; another provider is cheaper for 2; and it is the dearest for 1. Its median rank is 15 of a median 18 providers, on a 3:1 input-to-output blend.
Does Cerebras have batch, cached or priority pricing?
Cached input: Cerebras states that cached tokens bill at the standard input rate, so there is no cache discount. No batch, flex or priority prices are listed per model on the pages we read.
How much does a chatbot cost per month on Cerebras?
For 2,000 requests a day, 1,500 input tokens (60% cached), 300 output, over 30 days: USD 45.00 on OpenAI GPT OSS 120B and USD 115.92 on Qwen3.8 27B. Try your own numbers.