Agent services

Unlock Agent and API services with digital payments.

Services

aeon7

AEON-7 Gemma-4 31B: uncensored, NVFP4-quantized (dFlash) chat/completions. OpenAI-compatible at /services/aeon7/v1/chat/completions with a 256K context window. Request model 'gemma4-aeon-uncensored' (or 'gemma4-fast' for the speed profile). Pay per-MTok via x402 (OBOL on Ethereum or USDC on Base) -- no signup, no API key.

http

Pay with 2 options

0.0001 OBOL · ethereum
0.03 USDC · base

Model

aeon-7/gemma-4-31b-uncensored-nvfp4-dflash

Endpointhttps://inference.v1337.org/services/aeon7

API docsAPI docs ↗·openapi.json ↗

conf-spark1

Confidential dstack inference on spark1 (experimental stub TEE). Model qwen3.5:9b. Base Sepolia USDC demo.

inference

Price

0.001 USDC/request

Network

base-sepolia

Model

qwen3.5:9b

Endpointhttps://inference.v1337.org/services/conf-spark1

API docsAPI docs ↗·openapi.json ↗

How x402 payments work↓

A request without payment returns HTTP 402 with the price and a payment recipe. The buyer (or their library) signs an off-chain authorization, retries the request with an X-PAYMENT header, and the seller's facilitator settles on-chain. No gas needed — the seller covers settlement.

Useful for: data feeds, AI inference, subscriptions, digital purchases, and agent-to-agent commerce.

Spec: x402.org · Python SDK: pypi.org/project/x402