What is Featherless
Featherless provides access to open-weight language models through an OpenAI-compatible API, with flat subscriptions for chat and agent use plus token-billed business access.
Individual plans
As of August 5, 2026, Chat costs US$25 per month and includes access to every supported model, including Kimi and GLM families, four concurrent requests and a 32,000-token context allowance. Featherless states that prompts and responses are not logged.
Agent Standard costs US$100 per month and supports models up to 229 billion parameters, eight concurrent requests, a 256,000-token context window and one sandbox. Agent Pro costs US$200 per month and extends that access to any model while retaining eight concurrent requests, the 256,000-token context window and a larger sandbox.
Business options
Business 229B Full Context costs US$100 per unit monthly and Business Max costs US$200 per unit monthly. Per-Request starts at US$25 in prepaid monthly credit, removes the model-size limit and supports up to 100 concurrent requests. Credits are deducted only for successful requests and do not expire. Published examples per million tokens include GPT OSS 120B at US$0.10 input and US$0.55 output, GLM 5.2 at US$1.39 and US$4.40, and MiniMax M3 at US$0.55 and US$2.20.
Official sources: Featherless plan guide, Featherless current plans and model access.