# BridgeNode — Full Agent Guide

> AI inference bridge for AI agents. No API keys, no registration, no subscriptions.
> Pay per request with Solana USDC via x402 (HTTP 402). Transaction fees are sponsored — the agent only needs USDC.

## Endpoints

- OpenAI-compatible API: https://bridgenode.cc/v1
- Models & prices (JSON): https://bridgenode.cc/v1/models
- MCP server (streamable-http): https://bridgenode.cc/mcp
- Agent install map: https://bridgenode.cc/llms.txt
- Skill file: https://bridgenode.cc/skill.md

## Models & Pricing

Prices are in USDC per token (6 decimals). Live prices: `GET https://bridgenode.cc/v1/models`.

| Model | Input / token | Output / token | Context window | Max output |
|---|---|---|---|---|
| `gpt-oss-20b` 🆓 | $0.00000000 | $0.00000000 | 131,072 | 32,768 |
| `gpt-oss-120b` 🆓 | $0.00000000 | $0.00000000 | 131,072 | 32,768 |
| `glm-4.7-flash` 🆓 | $0.00000000 | $0.00000000 | 131,072 | 8,192 |
| `glm-4.5-flash` 🆓 | $0.00000000 | $0.00000000 | 131,072 | 8,192 |
| `glm-4.6v-flash` 🆓 | $0.00000000 | $0.00000000 | 131,072 | 8,192 |
| `deepseek-v4-flash` | $0.00000026 | $0.00000077 | 1,048,576 | 8,192 |
| `glm-4.7-flashx` | $0.00000008 | $0.00000047 | 1,048,576 | 8,192 |
| `glm-5.2` | $0.00000164 | $0.00000515 | 1,048,576 | 8,192 |
| `glm-5.1` | $0.00000164 | $0.00000515 | 1,048,576 | 8,192 |
| `glm-5` | $0.00000117 | $0.00000374 | 1,048,576 | 8,192 |
| `glm-5-turbo` | $0.00000140 | $0.00000468 | 1,048,576 | 8,192 |
| `glm-4.7` | $0.00000070 | $0.00000257 | 1,048,576 | 8,192 |
| `glm-4.6` | $0.00000070 | $0.00000257 | 1,048,576 | 8,192 |
| `glm-4.5` | $0.00000070 | $0.00000257 | 1,048,576 | 8,192 |
| `glm-4.5-x` | $0.00000257 | $0.00001041 | 1,048,576 | 8,192 |
| `glm-4.5-air` | $0.00000023 | $0.00000129 | 1,048,576 | 8,192 |
| `glm-4.5-airx` | $0.00000129 | $0.00000527 | 1,048,576 | 8,192 |
| `glm-4-32b-0414-128k` | $0.00000012 | $0.00000012 | 131,072 | 8,192 |
| `glm-5v-turbo` | $0.00000140 | $0.00000468 | 1,048,576 | 8,192 |
| `glm-4.6v` | $0.00000035 | $0.00000105 | 1,048,576 | 8,192 |
| `glm-4.6v-flashx` | $0.00000005 | $0.00000047 | 1,048,576 | 8,192 |
| `glm-4.5v` | $0.00000070 | $0.00000211 | 1,048,576 | 8,192 |
| `glm-ocr` | $0.00000004 | $0.00000004 | 131,072 | 8,192 |
| `kimi-k2.7-code` | $0.00000111 | $0.00000468 | 262,144 | 32,768 |
| `kimi-k2.7-code-highspeed` | $0.00000222 | $0.00000936 | 262,144 | 32,768 |
| `kimi-k2.6` | $0.00000111 | $0.00000468 | 262,144 | 32,768 |
| `kimi-k2.5` | $0.00000070 | $0.00000351 | 262,144 | 32,768 |
| `MiniMax-M2.7` | $0.00000035 | $0.00000140 | 1,048,576 | 8,192 |
| `MiniMax-M2.7-highspeed` | $0.00000070 | $0.00000281 | 1,048,576 | 8,192 |
| `MiniMax-M2.5` | $0.00000035 | $0.00000140 | 1,048,576 | 8,192 |
| `MiniMax-M2.5-highspeed` | $0.00000070 | $0.00000281 | 1,048,576 | 8,192 |
| `MiniMax-M2.1` | $0.00000035 | $0.00000140 | 1,048,576 | 8,192 |
| `MiniMax-M2.1-highspeed` | $0.00000070 | $0.00000281 | 1,048,576 | 8,192 |
| `MiniMax-M2` | $0.00000035 | $0.00000140 | 1,048,576 | 8,192 |
| `deepseek-v4-pro` | $0.00000077 | $0.00000232 | 1,048,576 | 8,192 |
| `kimi-k3` | $0.00000351 | $0.00001755 | 1,048,576 | 32,768 |
| `glm-5.3` | $0.00000164 | $0.00000515 | 1,048,576 | 8,192 |
| `minimax-m3` | $0.00000035 | $0.00000140 | 1,048,576 | 8,192 |

Pricing model: **exact scheme** — the agent pays for `input tokens + max_tokens` **before** processing. Minimum charge per request: 2000 atomic units = $0.002 USDC.

## Payment Flow (x402 V2, exact scheme)

1. Send the request without payment headers.
2. Server responds `402 Payment Required` with a `PAYMENT-REQUIRED` header (base64 JSON): price, `payTo` address, USDC mint, memo, recent blockhash.
3. Agent constructs a **partial transaction**: USDC `TransferChecked` (amount = required) + Memo instruction, signs with its own wallet. Fee payer is NOT signed by the agent.
4. Agent retries the request with `PAYMENT-SIGNATURE` header (base64 JSON payload with the signed transaction).
5. Server verifies the payment and processes the request (fees sponsored — gasless for the agent).
6. Response is `200` with `PAYMENT-RESPONSE` header (settlement receipt).

- Network: `solana:5eykt4UsFv8P8NJdTREpY1vzqKqZKvdp` (Solana mainnet)
- Asset: USDC `EPjFWdd5AufqSSqeM2qN1xzybapC8G4wEGGkZwyTDt1v`
- The agent must have an existing USDC ATA; it does **not** need SOL (BridgeNode sponsors fees).
- Use official x402 SDKs (`@x402/svm`, `x402[svm]`) or any x402-capable client — they handle 402 → sign → retry automatically.

## Quick Start (curl)

```bash
curl https://bridgenode.cc/v1/chat/completions   -H "Content-Type: application/json"   -d '{"model":"deepseek-v4-flash","messages":[{"role":"user","content":"hello"}],"max_tokens":100}'
```

Response: `402` with `PAYMENT-REQUIRED` header. Sign the partial transaction with an x402-capable client and retry with `PAYMENT-SIGNATURE` header. Response: `200` with the completion.

## SDKs

- Python SDK: `pip install bridgenode-llm` (https://pypi.org/project/bridgenode-llm)
- Python full toolkit (SDK + CLI): `pip install bridgenode` (https://pypi.org/project/bridgenode)
- CLI: `pip install bridgenode-cli` (https://pypi.org/project/bridgenode-cli)
- Alias packages (same toolkit): `bridgenode-sdk`, `bridgenode-mcp`, `bridgenode-skill` (https://pypi.org/project/bridgenode-sdk)
- TypeScript: `npm i @bridgenode/llm` (https://www.npmjs.com/package/@bridgenode/llm)
- MCP wrapper: `npm i @bridgenode/mcp` (https://www.npmjs.com/package/@bridgenode/mcp)

All SDKs handle the x402 payment handshake automatically, with fail-closed spending limits (`BRIDGENODE_MAX_PER_CALL`, `BRIDGENODE_DAILY_CAP`).

## MCP Usage

- One-line install: `claude mcp add bridgenode -s user -- npx -y @bridgenode/mcp@latest`
- Server URL: `https://bridgenode.cc/mcp` (streamable-http)
- Tool: `chat_completions` (model, mode, messages, max_tokens)
- Payment: x402 handshake per tool call; always check the actual amount in the 402 response before signing.

## Errors

| Status | Meaning |
|---|---|
| 400 | Bad request (unknown model, invalid body, oversized non-stream max_tokens) |
| 402 | Payment required — see `PAYMENT-REQUIRED` header |
| 413 | Request body too large (limit 2 MB) |
| 429 | Too many requests (queue limit) |
| 503 | Service busy — retry with backoff |

## Discovery

- Agent card: https://bridgenode.cc/.well-known/agent-card.json
- MCP manifest: https://bridgenode.cc/.well-known/mcp.json
- AI manifest: https://bridgenode.cc/.well-known/ai-manifest.json
- API catalog: https://bridgenode.cc/.well-known/api-catalog
- Listed on x402-list: https://x402-list.com/services/bridgenode
- Listed on x402-dev: https://www.x402dev.com/awesome-projects/
- Listed on nohumans.directory: https://nohumans.directory/l/f1f74751-9d5
- Listed on gold-402: https://github.com/Haustorium12/gold-402/blob/main/directory/learning.md
- ClawHub skill: https://clawhub.ai/bridgenode/skills/bridgenode

## Notes

- Refunds: if the provider fails before any content is delivered, the payment is refunded automatically (reverse USDC transfer).
- Reasoning models: use `max_tokens >= 200` — a too-small limit can produce an EMPTY answer; empty answers are NOT refunded (service provided).