/i — agent index

Machine-readable pricing, hosting, training, payment rails, and routing for image-edit inference. Start at /llms.txt or /i/index.json. Updated 2026-05-28. No auth on read endpoints. 2% routed + standard fees apply.

Start here
curl https://infr.sh/llms.txt
curl https://infr.sh/i/index.json
curl https://infr.sh/i/prices.json
SDK
npm install infr
import { Infr } from 'infr'
const infr = new Infr({ baseUrl: 'https://infr.sh' })
await infr.index() // → /i/index.json
await infr.prices.list()
Endpoints · 2% routed + standard fees apply
manifest Index manifest

Machine-readable map of all agent endpoints on infr.sh.

llms LLMs.txt

Standard agent discovery file (also at /llms.txt).

prices Managed inference pricing

Arena.ai leaderboard models + provider prices, normalized to $/image @1MP.

hosting GPU hosting & serverless

H100/A100/B200 rental and per-second serverless GPU rates for self-hosting.

training Training costs

Managed LoRA trainers + trainable open-weight models from the catalog.

rails Payment rails

Inbound/outbound rails (Tempo, x402, Stripe, onramps) with fees and limits.

onramps Onramps & 01.cash stack

Widest inbound coverage — Peer ZKP2P (Venmo, Revolut, Wise, Cash App), card onramps, stablecoins. anything → 01.cash → Pute → compute.

sharing Credit sharing

Which wallet/credit systems support P2P, gifting, and team pools.

routing Inference routing

How to call /api/v/{model} — auth paths, 2% fee, MPP/402 agents.

orchestrate Chain planner

LLM endpoint planner — primary chain + two alternatives at your volume.

quote Chain quote

Deterministic multi-step price quote from model IDs.

Human UI: prices · gpu · train · pay · docs