Integrations › MCP server

Connect Lenz to your coding tool.

Lenz is a remote Model Context Protocol server. Add it to Claude Code, Codex, Cursor, or any MCP client, and your assistant can fact-check a claim right inside the session — a quick verdict first, and a sourced verdict with citations from the deep check.

Nothing to install. It's hosted at lenz.io/mcp over Streamable HTTP, and it authenticates with the same lenz_… API key as the REST API — so there's no separate signup.

Using Claude Desktop, claude.ai or ChatGPT rather than a coding tool? Connect Lenz to Claude or ChatGPT → — the same server, as click-paths, with no config to edit.

Connect Lenz → View on GitHub API reference

Add the server

Point your client at https://lenz.io/mcp with a Bearer header. Set LENZ_API_KEY in your environment first, or have the config written for your client with the key already in it — Generate setup instructions →

Any MCP client .mcp.json

// add to your client's MCP config
{
  "mcpServers": {
    "lenz": {
      "type": "http",
      "url": "https://lenz.io/mcp",
      "headers": {
        "Authorization": "Bearer ${LENZ_API_KEY}"
      }
    }
  }
}

Claude Code one command

# register the remote server
$ claude mcp add --transport http lenz \
    https://lenz.io/mcp \
    --header "Authorization: Bearer ${LENZ_API_KEY}"

# then, in any session:
> verify: "The Great Wall is visible from space"

Codex one command

# registers the server in ~/.codex/config.toml:
$ codex mcp add lenz \
    --url https://lenz.io/mcp \
    --bearer-token-env-var LENZ_API_KEY

# no key enters the file — export it instead:
$ export LENZ_API_KEY=lenz_…

Full setup notes, per-client quirks, and examples live in the lenz-mcp repo → — Lenz is also listed in the Claude connectors directory ↗

What your assistant can do

Seven tools, same depth ladder as the API. Ask in plain language; the client picks the tool.

  • assess_claim — a quick verdict and a confidence for every claim in a text or a list, from a 3-model panel. About 15–20 seconds, one credit per claim, no sources shown.
  • verify_claim — investigates one claim through the full multi-model pipeline (research, debate, panel review): verdict, score, key finding, warnings and the main sources, in about a minute to a minute and a half. depth: low researches fewer sources for half the credits.
  • get_verification — wait for a running verify_claim by task id, or fetch a completed result by verification id.
  • ask_followup — a grounded follow-up question about a completed verify_claim, answered from its full evidence.
  • select_claims — choose which claim(s) to run when a text contains several.
  • check_usage — remaining credits, the per-tool price list, and your plan.
  • list_verifications — your recent deep checks, newest first. Use it to get back a result that finished after the chat moved on.

These are tools, not slash commands. In a client like Claude Code, just ask in plain language — “assess whether the earth is flat with Lenz” — and the assistant calls the right tool.

Prefer to call it in code instead? The same primitives are a REST API and typed Python / TypeScript SDKs — see the developer docs →

Pricing

Start free, no card required. The MCP server draws on the same credit pool as your API key — same key, same balance.

Free

$0

100 credits/month
enough for 100 /assess, or 10 /verify, or 100 /ask
/extract free · 1,000/day

Prototype today, zero spend. No card required.

Pro

$99/mo

5,000 credits/month
enough for 5,000 /assess, or 500 /verify, or 5,000 /ask
/extract free · 1,000/day

Self-serve. Annual plan: $83/mo, billed $990 yearly. For teams shipping AI-drafted text.

Scale

$399/mo

20,000 credits/month
enough for 20,000 /assess, or 2,000 /verify, or 20,000 /ask
/extract free · 1,000/day

Annual plan: $333/mo, billed $3,990 yearly. For high-volume production workflows.

Compare all plans on the plans page →

Enterprise

Volume beyond the Scale plan, SLAs, white-label, custom integration support.

Talk to us →

Frequently asked questions

A remote Model Context Protocol server hosted at lenz.io/mcp. It exposes Lenz's fact-checking pipeline as MCP tools, so an AI client like Claude Code, Codex, or Cursor can verify a claim against independent sources and get a quick verdict, and a sourced one from the deep check — without leaving the session. Claude Desktop, claude.ai and ChatGPT reach the same server through their own connector screens: connect Lenz to Claude or ChatGPT.

No. The server is remote, over Streamable HTTP. Point your MCP client at https://lenz.io/mcp with an Authorization: Bearer header carrying your API key. Get a free key at /api-credentials.

Seven: assess_claim (a quick 3-model verdict with a confidence, no sources shown), verify_claim (the full multi-model pipeline for one claim, with its main sources), get_verification (wait for a running verification, or fetch a completed one by id), ask_followup (grounded follow-up on a completed verification), select_claims (choose which claim to run when a text contains several), check_usage (remaining credits and plan), and list_verifications (your recent deep checks, newest first, to get back a result that finished after the chat moved on).

Yes. The MCP server is a thin proxy over the same public API at lenz.io/api/v1, authenticated with the same lenz_… Bearer key from /api-credentials. No separate signup.

Other ways to connect Lenz

Resources

lenz-mcp on GitHub Claude connectors directory Lenz in Claude or ChatGPT All integrations Developer docs API reference Get an API key API terms

Give your assistant a fact-checker.

Self-serve from day one. Free tier, no card required.

Connect Lenz →