CapCut Codex plugin plus Sume hosted MCP in one session
CapCut's Codex plugin drafts and edits in chat; Sume's hosted MCP at mcp.sume.com generates media. Here is what each does and how files pass between them.

You can run both in one Codex task as two separate tools. CapCut's plugin is a CapCut product; Sume is a remote MCP server at https://mcp.sume.com/mcp, which the docs say works in Codex among other clients. Neither page documents an integration between them, so plan on moving anything from Sume to CapCut as a file you hand over.
CapCut's page says the plugin is "now available outside the U.S." with a U.S. launch coming soon, and that it brings generation, drafting and detailed editing into a chat workflow. Sume facts are from Sume hosted MCP docs and MCP tools and gates, read 2026-09-30. The pages I read do not describe a connection between the two.
What does each side do?
| CapCut x Codex plugin | Sume hosted MCP | |
|---|---|---|
| What it is | CapCut plugin installed in Codex | Remote MCP server at mcp.sume.com |
| Described work | Generate or upload media, draft edits, refine in chat | Generation tools such as generate_video, music_create, tts_create |
| Availability | Outside the U.S.; U.S. launch coming soon | Hosted endpoint, OAuth or API key |
| Where results live | CapCut project (per CapCut's page) | media.sume.com artifacts and assets in your workspace |
How do Sume results get into the editor?
Ask Sume for the asset's download URL with assets_download_url (a read tool), fetch the file in your session, and give it to the CapCut plugin. CapCut's page says you can generate images and videos or upload an existing one, so an upload is the path CapCut's page describes. Whether the plugin can take a URL directly is not stated there, so check before relying on it.
Why can't Sume read my local clip?
The docs say "Hosted MCP cannot read files from your laptop." The route for a local file is: create an upload URL, have the client PUT the bytes, then call assets_complete. Our guide on uploading a local file over MCP walks through it. If CapCut produced the edit, upload the export that way before asking Sume to caption or extend it.
What happens when the agent calls a paid tool?
Paid writes such as generate_video need an idempotency_key. Long video jobs take time, so wait with jobs_wait rather than resubmitting; see MCP jobs_wait for long video jobs. For the broader question see can Codex make videos.
Sources
Related posts
More in Integrations
- Copilot mcp-config.json example: a Sume remote entry
A Copilot mcp-config.json example for Sume: an http entry at mcp.sume.com/mcp with an x-api-key header and a tools list. Global file vs workspace .mcp.json.
- Hedra MCP for Claude, and how Sume's hosted MCP connects
Hedra's MCP and CLI let Claude and other agents generate images, video and audio. Sume's hosted MCP lives at mcp.sume.com/mcp with write and paid gates.
- HeyGen translation SRT captions vs Sume burned-in captions
HeyGen now generates caption sidecars for every translation and lipsync job. Sume burns captions into the video and does not accept SRT as input.
- Instagram Audio API for Reels ads: prepare the video side
Meta's Instagram Audio API can find royalty-free replacements for copyrighted Reels music. Sume does not call it, but can drop the audio from your video.
Written by Sume