{
  "markdown": "# GPU-Bridge MCP Server\n\n> **30 GPU-powered AI services as MCP tools** — LLMs, image generation, audio, video, embeddings, reranking, PDF parsing, NSFW detection & more.\n> **x402 native** for autonomous AI agents: pay per request on-chain with USDC on Base L2. No API keys. No accounts.\n\n[![npm version](https://img.shields.io/npm/v/@gpu-bridge/mcp-server.svg)](https://www.npmjs.com/package/@gpu-bridge/mcp-server)\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![MCP Compatible](https://img.shields.io/badge/MCP-Compatible-blue.svg)](https://modelcontextprotocol.io)\n\n## What is GPU-Bridge?\n\n[GPU-Bridge](https://gpubridge.io) is a unified GPU inference API with **native x402 support** — the open payment protocol that allows AI agents to autonomously pay for compute with USDC on Base L2. No API keys, no accounts, no human intervention required.\n\nThis MCP server exposes all 30 GPU-Bridge services as Model Context Protocol tools, giving Claude (and any MCP-compatible AI) direct access to GPU inference.\n\n---\n\n## Install in Claude Desktop (2 minutes)\n\n### 1. Get your API key (or use x402 for autonomous agents)\n\nVisit [gpubridge.io](https://gpubridge.io) and grab a free API key, or use the x402 protocol for keyless agent payments.\n\n### 2. Add to `claude_desktop_config.json`\n\n**macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`\n**Windows:** `%APPDATA%\\Claude\\claude_desktop_config.json`\n\n```json\n{\n  \"mcpServers\": {\n    \"gpu-bridge\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"@gpu-bridge/mcp-server\"],\n      \"env\": {\n        \"GPUBRIDGE_API_KEY\": \"your_api_key_here\"\n      }\n    }\n  }\n}\n```\n\n### 3. Restart Claude Desktop\n\nThat's it. Claude now has access to 30 GPU-powered AI services.\n\n---\n\n## MCP Tools\n\n### `gpu_run`\nRun any GPU-Bridge service. The primary tool for executing AI tasks.\n\n```\nParameters:\n  service  (string)  — Service key (e.g., \"llm-4090\", \"flux-schnell\", \"whisper-l4\")\n  input    (object)  — Service-specific input parameters\n  priority (string)  — Optional: \"fast\" (lowest latency) or \"cheap\" (lowest cost)\n```\n\n### `gpu_catalog`\nGet the full catalog of available services with pricing and capabilities.\n\n### `gpu_estimate`\nEstimate cost before running a service. No authentication required.\n\n### `gpu_status`\nCheck the status of a job and retrieve results.\n\n### `gpu_balance`\nCheck your current balance, daily spend, and volume discount tier.\n\n---\n\n## 30 Available Services\n\n### Language Models (LLMs)\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `llm-4090` | General purpose LLM | Sub-second via Groq |\n| `llm-a100` | Maximum capability LLM | Largest models |\n| `llm-l4` | Ultra-fast, low cost LLM | Budget option |\n| `code-4090` | Code generation | Optimized for code |\n| `llm-stream` | Streaming LLM responses | Real-time output |\n\n### Image Generation\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `flux-schnell` | FLUX.1 Schnell | Fast, 4-step generation |\n| `flux-dev` | FLUX.1 Dev | High quality |\n| `sdxl-4090` | Stable Diffusion XL | Versatile |\n| `sd35-l4` | Stable Diffusion 3.5 | Latest SD model |\n| `img2img-4090` | Image-to-image | Style transfer, editing |\n\n### Vision & Image Analysis\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `llava-4090` | Visual Q&A | Image understanding |\n| `ocr-l4` | Text extraction (OCR) | Multi-language |\n| `rembg-l4` | Background removal | Instant |\n| `caption-4090` | Image captioning | Auto-describe images |\n| `nsfw-detect` | Content moderation | NSFW classification |\n\n### Speech-to-Text\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `whisper-l4` | Fast transcription | Sub-second |\n| `whisper-a100` | High accuracy transcription | Large files |\n| `diarize-l4` | Speaker diarization | Who said what |\n\n### Text-to-Speech\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `tts-l4` | Voice cloning TTS | 40+ voices |\n| `tts-fast` | Ultra-fast TTS | Lowest latency |\n| `bark-4090` | Expressive TTS | Emotion, laughter |\n\n### Audio Generation\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `musicgen-l4` | Music generation | Text-to-music |\n| `audiogen-l4` | Sound effects | Text-to-SFX |\n\n### Embeddings & Search\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `embed-l4` | Text embeddings | Multilingual |\n| `embed-code` | Code embeddings | For code search |\n| `rerank` | Document reranking | Jina, sub-second |\n\n### Video\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `animatediff` | Text-to-video | AnimateDiff |\n| `video-enhance` | Video upscaling | Up to 4K |\n\n### Utilities\n| Service ID | Description | Notes |\n|-----------|-------------|-------|\n| `pdf-parse` | Document parsing | PDF/DOCX to text |\n\n---\n\n## x402: For Autonomous AI Agents\n\nGPU-Bridge supports the [x402 payment protocol](https://x402.org), enabling truly autonomous AI agents to pay for compute without human intervention.\n\n```\nAgent Request → GPU-Bridge returns HTTP 402 Payment Required\n      ↓\nAgent pays USDC on Base L2 (gas < $0.01, settles in 2s)\n      ↓\nAgent retries with payment proof → GPU-Bridge executes and returns result\n```\n\n### Python Example with x402\n\n```python\nfrom x402.client import PaymentClient\n\nclient = PaymentClient(private_key=\"0x...\", chain=\"base\")\n\nresponse = client.request(\n    \"POST\",\n    \"https://api.gpubridge.io/v1/run\",\n    json={\n        \"service\": \"flux-schnell\",\n        \"input\": {\"prompt\": \"A robot painting on a canvas\", \"steps\": 4}\n    }\n)\nprint(response.json())\n```\n\n---\n\n## Pricing\n\n| Category | Starting From |\n|----------|--------------|\n| LLMs | $0.003/1K tokens |\n| Image Generation | $0.01/image |\n| Speech-to-Text | $0.005/minute |\n| Text-to-Speech | $0.005/1K chars |\n| Embeddings | $0.0001/1K tokens |\n| Reranking | $0.001/query |\n| PDF Parsing | $0.005/document |\n\nAll prices in USD. x402 payments in USDC on Base L2.\n\n---\n\n## Links\n\n- **Website:** [gpubridge.io](https://gpubridge.io)\n- **Docs:** [gpubridge.io/docs](https://gpubridge.io/docs)\n- **Catalog:** [gpubridge.io/catalog](https://gpubridge.io/catalog)\n- **GitHub:** [github.com/gpu-bridge](https://github.com/gpu-bridge)\n- **npm:** [@gpu-bridge/mcp-server](https://www.npmjs.com/package/@gpu-bridge/mcp-server)\n\n---\n\n## License\n\nMIT © [Healthtech Capital LLC](https://gpubridge.io)\n",
  "bytes": 6470,
  "sha": "afa75d6779d118bbbb86d222af8e38423484c9f303ed44cc8f4e7c20bb034f5d",
  "repo_slug": "gpu-bridge/mcp-server",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_fjnunezp75_gpu_bridge_6fb3915d/readme"
}