Featherless LLM
Serverless hosting for 40,000+ open-source LLMs via one API with flat, unlimited-token pricing from $25/mo.
What it does
Featherless is a serverless LLM hosting provider that gives developers instant access to 40,000+ open-source models from Hugging Face through a single API key. It offers flat monthly pricing with unlimited tokens and adds agent runtimes with sandboxes on higher tiers.
How to use: Users can access and run LLMs through the Featherless API. Business customers can deploy models through their dashboard, while individual plan users can request models via Discord or email.
Core features
One API for 40,000+ open-source models
Serverless inference with no setup or hosting
Flat pricing with unlimited tokens
Agent runtimes with sandbox and storage
Support for large models and long context
Best for
→Testing many open-source LLMs quickly
→Running production inference at flat cost
→Building agents on open models
Pricing
Premium
$25/mo
Agent Standard
$100/mo
Agent Pro
$200/mo
Toolspool rankingby monthly traffic