Start Here
What is camelStream?
Unlimited frontier intelligence at a flat monthly price, through OpenAI- and Anthropic-compatible APIs
camelStream provides one endpoint with OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages compatibility, serving unlimited frontier intelligence from a fleet of models above a published intelligence floor. Requests beyond your concurrent stream allowance queue automatically instead of failing.
How pricing works
A subscription covers the number of concurrent streams you choose, from 1 to 50, at a flat monthly price per stream. Tokens are unlimited while the subscription is active, and you can cancel anytime. Change your stream count in Stream Settings; Stripe prorates the change for the current billing period. See stream.camelai.com for current pricing, and contact us if you need more than 50 streams.
| Base URL | https://stream.camelai.com/v1 |
| Model ID | auto |
The model ID auto means your request is served by whatever fleet model is
available. See the fleet.
Get started
The models serving streams, the intelligence floor, and the guarantees.
Create a key and stream your first response.
Point Hermes, Claude Code, Codex, OpenCode, or Pi at Stream.
Explore the supported OpenAI and Anthropic endpoints.
Use existing OpenAI and Anthropic SDKs.
Understand concurrency, queues, and error responses.
Looking for the camelAI coding agent? These pages cover camelStream, the inference API. The agent that builds and deploys apps from a chat lives in the Coding Agent docs.