APA://SERVICES/LLAMA-GPT-INFERENCE

LLaMA GPT Inference

Metered LLaMA-70B inference billed per output token with machine-readable pricing headers.

ARC/ USDCDEMO MODE

← MARKETPLACE

PRICE

0.004 USDC

PER CALL

LATENCY

640 ms

DEMO FIXTURE

AVAILABILITY

99.85%

TIER PREMIUM

NETWORK

ARC · USDC

CHANNEL

402 PAYMENT FLOWAVAILABLE

1402 PAYMENT REQUIRED
Service responds 402 Payment Required with price + nonce.
2PAYMENT AUTHORIZED
Agent's wallet signs EIP-191 payment authorization.
3PAYMENT VERIFIED
Signature verified + provider executes API call.
4SERVICE RESPONSE
Settled receipt + service payload delivered.

APA://SERVICES/llama-gpt-inference/SESSION

SIMULATION

idle — click PAY & EXECUTE to request this endpoint.

{
  "output": "Agentic payments in 2026 have matured into three pillars: (1) nanopayment channels settling sub-cent API calls, (2) policy-driven agent wallets enforcing per-transaction ceilings, and (3) oracle-attested receipts enabling downstream verification. USDC on ARC has emerged as the dominant settlement rail with ~1.2B monthly agent-to-agent transfers.",
  "tokens_in": 42,
  "tokens_out": 89,
  "latency_ms": 624,
  "cost_breakdown": { "input": 0.00084, "output": 0.00356 },
  "settlement_tx": "0x4a...1f2d"
}

SERVICE

LLaMA GPT Inference

Metered LLaMA-70B inference billed per output token with machine-readable pricing headers.

PROVIDER

inference.arc