TensorFeed.ai

TensorFeed.ai is agent fair-trade certified: open pricing, automatic no-charge on 5xx, breaker, schema fail, and stale data, Ed25519-signed receipts on every paid call, inference-only license. Built with Claude (Anthropic). Standard at /.well-known/agent-fair-trade.json.

Network of two. TensorFeed.ai and TerminalFeed.io are AFTA-federated as of 2026-04-30. A single bearer token works on both sites via the cross-Worker validate + commit rail. Two independent Ed25519 receipt keypairs, single credit ledger, one open standard. Other publishers can self-adopt by publishing a conforming manifest.

x402 + Stripe Link. TensorFeed has been x402-compliant since the agent payments rail shipped. Today we accept the exact scheme with USDC on Base mainnet. Stripe Link (released April 2026) uses x402 with method=stripe for Shared Payment Tokens; supporting that method as a parallel x402 path is under evaluation. The protocol is the same; we already speak it.

Real-time AI industry intelligence for humans and AI agents. News from 15+ sources, live service status for every major LLM provider, model pricing and benchmark history, an AI agents directory, and a pay-per-call premium API settled in USDC on Base mainnet (no accounts, no API keys).

Site: https://tensorfeed.ai

Built with Claude: TensorFeed was designed by Ripper in collaboration with Claude (Anthropic). Specific systems Claude designed alongside: the agent payments rail, the active LLM probes, the GPU pricing aggregator, the OFAC sanctions screening pipeline, the routing engine, and the Agent Fair-Trade Agreement (AFTA) standard itself. We credit the collaboration openly, and the git log shows the build trail.

What this repo contains

• Web dashboard (src/): Next.js 14 App Router, static export, deployed to Cloudflare Pages
• API backend (worker/): Cloudflare Worker tensorfeed-api, attached to tensorfeed.ai/api/*. Cron-driven RSS polling, status scraping, model catalog updates, daily history snapshots, daily MCP server registry telemetry
• MCP server (mcp-server/): Model Context Protocol server published as @tensorfeed/mcp-server on npm. Exposes 19 tools (6 free, 13 paid) for use in Claude Desktop, Claude Code, Cursor, Cline, and any other MCP client. See mcp-server/README.md for the full tool list and setup
• Python SDK (sdk/python/): pip install tensorfeed. Full coverage of free + premium endpoints, optional [web3] extra for one-call USDC sign-and-send
• JavaScript SDK (sdk/javascript/): npm install tensorfeed. Same surface as the Python client

MCP server quick start

The TensorFeed MCP server is the fastest way to plug an AI agent into the TensorFeed API.

npx -y @tensorfeed/mcp-server

Or in your Claude Desktop / Claude Code config:

{
  "mcpServers": {
    "tensorfeed": {
      "command": "npx",
      "args": ["-y", "@tensorfeed/mcp-server"]
    }
  }
}

Free tools work without auth. To enable the 13 paid tools (routing recommendations, news search, history series, cost projection, provider deep-dive, model comparison, MCP registry series, webhook watches, and more), add a TENSORFEED_TOKEN env var. Buy credits at tensorfeed.ai/developers/agent-payments using USDC on Base. New wallets get a 50-credit welcome bonus on first payment.

Listed in the official MCP server registry as ai.tensorfeed/mcp-server. See mcp-server/README.md for the full tool catalog, example agent queries, and pricing.

API

All endpoints live at https://tensorfeed.ai/api/*.

Free, no auth:

• /api/news, /api/status, /api/models, /api/benchmarks, /api/incidents
• /api/agents/activity, /api/agents/news, /api/agents/status, /api/agents/pricing, /api/agents/directory
• /api/podcasts, /api/trending-repos
• /api/history, /api/history/{YYYY-MM-DD}/{type} (daily snapshots)
• /api/mcp/registry/snapshot (daily telemetry of the official MCP server registry)
• /api/health, /api/ping, /api/meta, /api/cron-status

Paid (1 credit each, USDC on Base):

• /api/premium/routing, /api/premium/news/search, /api/premium/cost/projection
• /api/premium/history/pricing/series, /api/premium/history/benchmarks/series, /api/premium/history/status/uptime
• /api/premium/agents/directory, /api/premium/providers/{name}, /api/premium/compare/models
• /api/premium/whats-new (agent morning brief), /api/premium/mcp/registry/series
• /api/premium/watches (webhook watches with daily/weekly digest tier)

Full documentation: tensorfeed.ai/developers and tensorfeed.ai/developers/agent-payments. Machine-readable: /llms.txt, /openapi.json, /.well-known/x402.json.

Discovery surfaces for AI agents

• Official MCP server registry: ai.tensorfeed/mcp-server
• llms.txt: https://tensorfeed.ai/llms.txt
• OpenAPI 3.1 spec: https://tensorfeed.ai/openapi.json
• x402 V2 manifest: https://tensorfeed.ai/.well-known/x402.json
• Agent-first landing: https://tensorfeed.ai/for-ai-agents
• View as Agent toggle: every page on the site exposes the underlying API JSON via a navbar toggle, so agents can introspect what powers any view

Stack

Next.js 14 (static export), Cloudflare Pages + Workers + KV, Tailwind CSS, JetBrains Mono + Inter, Resend for email alerts, Cloudflare Web Analytics. The MCP server is plain TypeScript on top of the official @modelcontextprotocol/sdk.

Development

npm install
npm run dev      # Next.js dev server at localhost:3000
npm run build    # Static export
npm run lint

Worker:

cd worker
npm install
npm test         # 214 vitest cases, all green
wrangler deploy

MCP server:

cd mcp-server
npm install
npm run build
npm start

License

MIT. See LICENSE.

Contact

• support@tensorfeed.ai
• press@tensorfeed.ai
• feedback@tensorfeed.ai

A Pizza Robot Studios project.