Telnyx x402 Gateway

Pay per request for Telnyx Inference.

OpenAI-compatible chat completions via Telnyx Inference. One endpoint, four models, pay-per-request in USDC. Agents can call the endpoint with a wallet only: no Telnyx account, no API key, and no subscription.

Endpoint POST /v1/chat/completions

Price $0.003–$0.10 USDC per call

Models (select via body.model)

MiniMaxAI/MiniMax-M2.7 ($0.003), moonshotai/Kimi-K2.6 ($0.02), zai-org/GLM-5.1-FP8 ($0.035), moonshotai/Kimi-K2.6-long ($0.10)

Payment x402 exact on Base mainnet (eip155:8453)

Send an unsigned request to receive the PAYMENT-REQUIRED challenge, then retry with a signed PAYMENT-SIGNATURE header using an x402 client. Use GET /v1/models to discover model ids; price varies by selected model.

curl -i -X POST https://x402.telnyx.com/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{"model":"MiniMaxAI/MiniMax-M2.7","messages":[{"role":"user","content":"Say hello in one short sentence."}],"max_tokens":64}'

# Discover available models (no payment):
curl https://x402.telnyx.com/v1/models

Payment schemes: exact, upto — upto supports EIP-2612 sponsored approval on USDC.

Discovery: GET /v1/models · GET /.well-known/x402.json

x402 clients can use the standard challenge from the response header. See the x402 protocol for client libraries.