Bridge by Wekoya / Guides / Use Bridge from your AI assistant with MCP

Use Bridge from your AI assistant with MCP

In short: Add Bridge's MCP server, https://wekoya-api.wekoya.tech/mcp, to your assistant. In Claude or ChatGPT you sign in and set a monthly naira cap; in Claude Code, Cursor, VS Code or Codex you use a Bridge key. The assistant can then compare models, estimate a feature's cost, find a cheaper model that passes your examples and explain your spending.

Updated 8 November 2026

What your assistant can do with it

• Answer "which model should I use and what will it cost?" from the live price list, in naira.
• Run one prompt on several models and show answers, cost and speed side by side, with a judge's score.
• Estimate a feature's monthly cost before anyone builds it.
• Find the cheapest model that still passes your own examples, quoted first, and put it live under a stable name.
• Explain what you spent, what changed and why, and what to change.
• Make a capped key for another agent, and set your users' budgets.

Connect it

1. Claude or ChatGPT: add a custom connector with the address https://wekoya-api.wekoya.tech/mcp, sign in to Bridge when asked, and choose the access, monthly cap and expiry.
2. Claude Code: claude mcp add --transport http bridge https://wekoya-api.wekoya.tech/mcp --header "Authorization: Bearer $WEKOYA_API_KEY".
3. Cursor, VS Code or Cline: add the server with the same address and an Authorization: Bearer header, as in the MCP docs.
4. Codex: add [mcp_servers.bridge] with the address and bearer_token_env_var = "WEKOYA_API_KEY" to ~/.codex/config.toml.

Then ask, for example: "Explain my AI spending this week" or "Find a cheaper model for our order extraction and show me the quote".

Read the MCP docs