Fable 5 at Opus pricing.
GPT‑5.5 at half price.
Frontier inference at half retail price, live on the Claw Hunter API and paid per call with x402. No API key, no subscription: your agent swaps a base URL and starts saving.
For agentsPaste this into your agent:
Read https://clawhunter.fun/llms.txt and use it for inference calls.
OpenAI-shaped: POST /api/v1/chat/completions · free model menu at /api/v1/chat/models
Claude Fable 5
anthropic
$5 / $25per 1M tokens, in / out
retail $10 / $50
$4.50 / $22.50 in $CLAWHUNTER, $SQUIRE, or $ANSEM
Anthropic's most capable model, served at Opus pricing.
GPT-5.5
openai
$2.50 / $15per 1M tokens, in / out
retail $5 / $30
$2.25 / $13.50 in $CLAWHUNTER, $SQUIRE, or $ANSEM
OpenAI's flagship at half of OpenAI's price.
Claude Opus 4.8 $2.50/$12.50Claude Sonnet 5 $2/$10Claude Sonnet 4.6 $1.50/$7.50GPT-5.4 mini $0.375/$2.25DeepSeek V4 Flash $0.07/$0.15GLM-5 $0.4/$1.20Kimi K2.6 $0.35/$1.75c0mpute Pro $0.1/reqc0mpute Max $0.15/reqc0mpute Max Think $0.2/reqmore models coming
Agents pay in $CLAWHUNTER, $SQUIRE, or $ANSEM and save another 10%
Every request quotes four rails: USDC, $CLAWHUNTER, $SQUIRE, and $ANSEM. The token rails are priced 10% under the USDC rate, so the cheapest way for an agent to run frontier inference is to spend the ecosystem's tokens.
any agent
Claw Hunter APIpaying in USDCstandard ratepaying in $CLAWHUNTERsave 10%paying in $SQUIREsave 10%paying in $ANSEMsave 10%
Try it yourself
This is the exact call agents make, at the same price. Pick a model, sign the x402 payment in your wallet, and get the completion with the receipt on-chain.
loading the demo…