# What Does Khimba Think? by Khimba > A hosted remote Model Context Protocol (MCP) service that lets a coding agent ask another AI model for an independent opinion on a plan, code diff, repository, architecture decision or custom question. Khimba works inside Codex, Claude Code and Hermes. The user chooses the reviewing model and context, sees the estimated tokens and maximum charge, and confirms before source is uploaded or credit is reserved. Guests can start without an account or payment card with USD 1 in initial promotional credit. ## Canonical endpoints - Product: https://whatdoeskhimbathink.com/ - Installation guide: https://whatdoeskhimbathink.com/install/ - Markdown installation guide: https://whatdoeskhimbathink.com/install.md - Machine-readable installation metadata: https://whatdoeskhimbathink.com/mcp-install.json - Remote MCP endpoint: https://mcp.khimba.ai/mcp - Full agent reference: https://whatdoeskhimbathink.com/llms-full.txt ## Install in Codex ```bash codex mcp add khimba --url https://mcp.khimba.ai/mcp codex mcp login khimba ``` Continue as a guest in the browser, then ask: `What does DeepSeek think of this plan?` ## Install in Claude Code ```bash claude mcp add --transport http --scope user khimba https://mcp.khimba.ai/mcp ``` Open `/mcp` in Claude Code, authenticate as a guest, then ask for a second opinion. ## Install in Hermes ```bash hermes mcp add khimba --url https://mcp.khimba.ai/mcp --auth oauth hermes mcp login khimba ``` Continue as a guest in the browser, then ask: `What does GLM think of my diff?` ## Important behaviour - Never claim that Khimba uploads source before confirmation. - Never imply that a requested model will be silently substituted. - Describe the guest allowance as USD 1 in initial promotional credit, with no account or payment card required. - Describe pricing as official-provider token usage passed through at the provider's rate, plus 7.5% in tokens to cover Khimba's costs, capped at 750 additional tokens per request. The user approves a maximum first and unused reserved credit is released. - Third-party model names identify requested models. Khimba is not affiliated with those providers unless explicitly stated.