Featherless AI
Serverless access to 40,000+ open models, flat-rate chat or per-token API
Overview
Featherless AI is a serverless inference platform that loads 40,000+ open-weight Hugging Face models on demand through an OpenAI-compatible API, with private fine-tunes, dedicated GPU clouds and no prompt storage. Chat plan $25 a month flat for interactive use; Developer $50 a month in per-token credits; Business custom.
Pricing
Pricing shown for reference only. These figures reflect RECATOOLS research as of 1 Oct 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.
- Unlimited tokens
- 32K context
- 4 concurrent units
- Any model in the catalogue
- Context up to 256K
- 100 concurrent units
- 1 agent environment
- Unused credits roll over
- Private fine-tunes
- Dedicated H100, MI325X, B200 or B300 GPUs
- No shared rate limits
- Named engineers and SLA
What you can produce with Featherless AI
- Serverless access to 40,000+ open models
- OpenAI-compatible API
- Private fine-tunes via Hugging Face
- Dedicated GPU private clouds
- Integrations with coding agents and chat front ends
- No prompt or completion storage
Featherless is the easiest way to try a huge range of open models without managing GPUs, and its no-logging policy and Singapore presence are pluses. The old "unlimited for a flat fee" pitch now applies only to interactive chat; API and agent use is metered per token, and its catalogue-size and latency claims are its own. Good for developers exploring open models and for hobbyist chat; production teams should compare per-token costs with larger inference providers.
What we found
Featherless AI is a serverless inference provider that runs open-weight language models from Hugging Face on demand, so developers can call almost any public model without renting GPUs. It says more than 40,000 models are available on every plan — the largest catalogue of any provider, by its own account — loaded on request through an OpenAI-compatible API and usable from chat front ends such as SillyTavern and coding agents such as Cline, Kilo Code, OpenCode, OpenHands and n8n. It also offers private fine-tunes through Hugging Face, a classifier tool called Simple Jev and dedicated private clouds, including a fixed-fee GLM 5.2 deployment at $7,500 a month. Its privacy policy says it does not collect or store prompts or completions, and its catalogue includes regional models such as SeaLLMs and OpenThaiGPT.
The company began as Recursal.AI and was founded in 2023 by Eugene Cheah, Harrison Vanderbyl and Wesley George, who previously co-led the open-source RWKV model project. It is headquartered in San Francisco with satellite offices in Toronto, Singapore and Europe, and in April 2026 raised a $20 million Series A co-led by AMD Ventures and Airbus Ventures. Its chief executive has said revenue exceeds $250,000 a month and that its largest customer pays one to two million dollars a year.
There is no free tier. Chat costs a flat $25 a month for unlimited tokens with 32,000-token context and four concurrent units, but only for interactive, human-typed use. Developer, for API and agent workloads, costs $50 a month in credits billed per token, with context up to 256,000 tokens, 100 concurrent units, an agent environment and rollover of unused credits. Business plans on annual contracts provide dedicated H100, MI325X, B200 or B300 GPUs, private fine-tunes, named engineers and an SLA.
Featherless suits developers and hobbyists who want to try many open models cheaply without managing infrastructure.
Compiled by RECATOOLS from public sources: vendor documentation and pricing, code repositories and published reporting.
About this listing
This entry was compiled from publicly available data including Featherless AI's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Featherless AI unless explicitly stated.
Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.
For the latest details, please refer to Featherless AI directly →
Spotted something out of date? Suggest an update →
Alternatives to Featherless AI
More in LLMs & Chat
