Eddie AI Adds MCP Access and Usage Pricing
Eddie AI added a Model Context Protocol server on August 27, enabling its video-editing assistant to run from Claude and ChatGPT through an OAuth-connected custom connector. No Film School reports that the company also introduced pay-as-you-go access and released an iPhone app. Eddie's documentation specifies the MCP endpoint and authentication flow for compatible clients.
Eddie AI has released a Model Context Protocol (MCP) server that lets users invoke its video-editing assistant from Claude or ChatGPT. The company's MCP documentation lists https://mcp.heyeddie.ai/api/mcp as the custom connector endpoint and specifies OAuth authentication, with no client secret or token required for users to paste.
No Film School reported on August 27 that Eddie AI also introduced pay-as-you-go access, rather than requiring a monthly fee, and released an iPhone app. The publication describes the assistant as supporting tasks including transcription, soundbite selection, assembly, b-roll, captions, music, voiceover, multicam work, and NLE export.
Connector implementation
Eddie's documentation instructs users to add a custom connector named "Eddie AI" in Claude or ChatGPT, then authenticate through an Eddie account. According to the documentation, an unauthenticated request receives a 401 response with a WWW-Authenticate header that identifies the OAuth protected-resource metadata endpoint. A spec-compliant MCP client can then register and prompt the user to sign in.
The documentation explicitly identifies Claude and ChatGPT as supported connection targets. It also notes that the desktop app is recommended for ChatGPT's add-MCP flow.
Editing workflow and implications
No Film School reports that, in ChatGPT, returned edits can appear as playable cards using frames from the user's footage rather than as external links. Its report also describes the iPhone workflow as accepting footage from the camera roll, Drive, or Frame.io; users can request an edit, review it in a gapless player, save an MP4, or share a link.
For AI-tool builders, this release is a concrete example of MCP being used to connect a domain-specific media service to general-purpose assistants. Comparable MCP integrations can reduce context switching by exposing specialized operations inside an existing chat interface, but practical use still depends on the client's connector support, OAuth implementation, permission boundaries, and how reliably the service handles media assets and export jobs.
The available material does not provide pricing rates, credit-unit definitions, supported video formats, processing-time benchmarks, or API-level details for the editing operations. Those details remain relevant for teams assessing operational cost and production integration.
Key Points
- 1Eddie AI's MCP server exposes video-editing workflows inside Claude and ChatGPT through a documented OAuth connector flow.
- 2No Film School reports pay-as-you-go access and an iPhone app, expanding available entry points for Eddie's editing service.
- 3Comparable MCP integrations can reduce tool switching, while OAuth, media handling, and export reliability remain key implementation considerations.
Scoring Rationale
This is a useful MCP integration for practitioners building or evaluating agent-accessible creative workflows. Its direct relevance is strongest for video-production tooling rather than the broader ML infrastructure ecosystem, and the available reporting lacks performance and pricing specifics.
Sources
Primary source and supporting public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
