# MapleAI — AI Agent Discovery ## Service Overview - **Name**: MapleAI - **Type**: LLM API provider (GPT models) - **Protocol**: x402 (pay-per-request on Arc) - **Base URL**: https://arc.mapleai.shop/v1 - **Compatibility**: OpenAI Chat Completions + OpenAI Responses API - **Contact**: contact@mapleai.shop ## Available Models 4 GPT models, priced per token. Full catalog: GET https://arc.mapleai.shop/v1/models - **openai/gpt-5.6-sol** — GPT-5.6 Sol, 1.1M context, $2.80 in / $14.00 out per 1M tokens - **openai/gpt-5.6-terra** — GPT-5.6 Terra, 1.1M context, $1.40 in / $8.40 out per 1M tokens - **openai/gpt-6-luna** — GPT-6 Luna, 1.1M context, $0.07 in / $0.35 out per 1M tokens - **openai/gpt-6-sol** — GPT-6 Sol, 1.1M context, $1.40 in / $7.00 out per 1M tokens ## Image Models - gpt-image-2 (1024x1024): $0.0200 per image - gpt-image-2.5 (1024x1024): $0.0400 per image - gpt-image-2.5-flare (1024x1024): $0.0400 per image - gpt-image-2.5-sunburst (1024x1024): $0.0400 per image - grok-imagine-image (1024x1024): $0.0450 per image - gpt-image-2-2k (2048x2048): $0.0900 per image - gpt-image-2.5-sunburst-2k (2048x2048): $0.2200 per image POST https://arc.mapleai.shop/api/v1/images/generations generates images. POST https://arc.mapleai.shop/api/v1/images/image2image edits a PNG, JPEG or WebP supplied as a base64 data URI (maximum 10 MB). Send model, size, prompt and optional n (1-4); edits also require image. The x402 challenge includes the exact price with payment overhead. Successful responses contain data[].url or data[].b64_json. ## Jev - jev-latest: $0.12 per 1M input tokens, output free, plus settlement overhead. POST https://arc.mapleai.shop/jev uses SystemOne; send model, state and named questions, each with type and instructions. Read answers from the response. ## Free Embeddings - nvidia/nemotron-3-embed-1b: free POST https://arc.mapleai.shop/v1/embeddings; send input as a string or array of strings. ## Free Chat (gpt-oss-20b) - nvidia/gpt-oss-20b: free POST https://arc.mapleai.shop/v1/free/chat/completions; no payment. Rate-limited 10/10min and 100/day per agent IP; max_tokens capped at 2048. Check remaining window: GET https://arc.mapleai.shop/v1/free/chat/completions/quota. ## Prepaid API Keys POST https://arc.mapleai.shop/prepaid/codes (x402-paid) issues a prepaid bearer key for one GPT model with a token budget in 100000-token steps from 100000 to 1000000, priced at the model input rate plus settlement fee. Use the key at https://mapleai.shop/v1 (OpenAI-compatible). Check usage and status for free: GET https://mapleai.shop/v1/prepaid/status with the prepaid key as the Bearer token — returns valid, reason and tokens total/used/reserved/remaining. Error recovery: an invalid purchase body is answered with availableModels and their pack prices; a prepaid chat call whose model is outside the key returns 403 with the current allowedModels of that key; an exhausted key returns valid=false with reason. In every case the response names https://mapleai.shop/v1/prepaid/status as the refresh URL before buying a new pack. ## Quick Start for AI Agents The code below shows the request format. A standard OpenAI SDK does not handle x402 payment automatically. Use an x402-aware client to read the 402 challenge, sign it and retry with PAYMENT-SIGNATURE. ### Python ```python from openai import OpenAI client = OpenAI( base_url="https://arc.mapleai.shop/v1", api_key="payment-signature-required" # use an x402-aware client ) response = client.chat.completions.create( model="openai/gpt-5.6-sol", messages=[{"role": "user", "content": "Hello"}] ) ``` ### TypeScript ```typescript import OpenAI from 'openai'; const client = new OpenAI({ baseURL: 'https://arc.mapleai.shop/v1', apiKey: 'payment-signature-required' // use an x402-aware client }); const response = await client.chat.completions.create({ model: 'openai/gpt-5.6-sol', messages: [{role: 'user', content: 'Hello'}] }); ``` ### Responses API ```bash curl https://arc.mapleai.shop/api/v1/responses \ -H 'content-type: application/json' \ -d '{"model":"openai/gpt-5.6-sol","input":"Hello"}' ``` ## Discovery Endpoints Developer guide with network URLs, model prices and request examples: https://arc.mapleai.shop/developers ### Service Index `GET https://arc.mapleai.shop/service-endpoints.json` One-shot machine-readable index: every route with its access mode (x402 / free / prepaid bearer), live per-unit pricing and request examples. ### x402 Resource Manifest `GET https://arc.mapleai.shop/.well-known/x402` Machine-readable x402 resource manifest for wallet/agent discovery. ### A2A Agent Card `GET https://arc.mapleai.shop/.well-known/agent-card.json` A2A-protocol agent card: skills, x402 security scheme and the OpenAI-compatible HTTP interface; use it for agent-to-agent service discovery. ### OpenAPI Specification `GET https://arc.mapleai.shop/openapi.json` Full API specification in OpenAPI 3.1 format, including per-model pricing. ### Model Catalog `GET https://arc.mapleai.shop/v1/models` Free list of available models with pricing and context windows. ## Paid Endpoints | Method | Path | Protocol | | --- | --- | --- | | POST | `/v1/chat/completions` | OpenAI Chat Completions | | POST | `/api/v1/chat/completions` | OpenAI Chat Completions (alias) | | POST | `/api/v1/responses` | OpenAI Responses API (alpha) | | POST | `/v1/responses` | OpenAI Responses API (alpha) | | POST | /api/v1/images/generations | Image generation | | POST | /api/v1/images/image2image | Image editing | | POST | /jev | Jev SystemOne decisions | | POST | /v1/free/chat/completions | Free gpt-oss-20b chat (quota-limited) | | POST | /prepaid/codes | Buy a prepaid API key | | GET | https://mapleai.shop/v1/prepaid/status | Prepaid key usage and status (free) | ## Payment Protocol 1. Send the request without payment 2. Receive HTTP 402 with the exact USDC amount in the `PAYMENT-REQUIRED` header 3. Sign the USDC transfer with your wallet 4. Retry with the `PAYMENT-SIGNATURE` header — the call executes immediately Price = counted input tokens × input rate + requested output tokens × output rate, floored at $0.0010 per paid request. Failed calls (HTTP >= 400 from the model provider) are cancelled, not settled. ## Key Features - **4 GPT models**: GPT-5.6 Sol, GPT-5.6 Terra, GPT-6 Luna and GPT-6 Sol - **Pricing**: 30% below published OpenAI rates, from $0.07 per 1M input tokens - **OpenAI-compatible request format**: requires an x402-aware payment client - **x402 payments**: pay per request in USDC on Arc - **No accounts**: your wallet address is your identity ## Technical Details - **Network**: eip155:5042 (Arc mainnet) - **Payment token**: USDC (0x3600000000000000000000000000000000000000) - **Treasury (payTo)**: 0x63db6eaf635a31bbc6714fe37bdc85243864f611 - **Facilitator**: http://127.0.0.1:4030 - **Minimum charge**: $0.0010 per paid request - **Max context**: 1.1M tokens ## Use Cases - Autonomous AI agents with budget control - Multi-agent systems (pay per agent, per request) - Research and prototyping with no commitment - Enterprise deployments that cannot hold API keys ## Support - API status: https://arc.mapleai.shop/health - Contact: contact@mapleai.shop - x402scan: https://x402scan.com - Protocol docs: https://x402.org --- **For AI agents**: every endpoint returns structured JSON. Point the OpenAI SDK at `https://arc.mapleai.shop/v1` with any non-empty `api_key` and handle the 402 challenge with an x402 client. Last updated: 2026-09-30.