Edit video in ChatGPT with Valmera
ChatGPT can't render video by itself, but once Valmera is added as a custom MCP server it can edit footage you upload. It cuts silences and filler words, adds captions, reframes to 9:16, adds music and b-roll, renders a preview and exports the full-quality MP4. You add the server once on ChatGPT's Plugins page and sign in to Valmera.
MCP SERVER URL
https://valmera.io/mcp/serverPrefer a guided version? The connect page shows the steps and confirms when Valmera receives ChatGPT's first request.
Add Valmera to ChatGPT
- 1Create a custom MCP serverOpen chatgpt.com/plugins, select +, then Create custom MCP server. If the option isn't there, turn on Developer mode in Settings → Security and login.
- 2Paste the URLName it Valmera, add a short description, and paste https://valmera.io/mcp/server under Connection.
- 3Keep OAuth and create itLeave authentication on OAuth. Valmera supports dynamic client registration, so there is no client ID or secret to enter. Review the risk notice, select I understand and want to continue, then Create as a plugin.
- 4Sign in to ValmeraSign in with the email or Google account you subscribed with and choose Allow connection. ChatGPT then lists Valmera's tools.
- 5Use it in a chatStart a new conversation, type @, choose Valmera and ask: List my Valmera projects. An empty list still means you're connected.
Steps follow OpenAI's connection guide, checked October 6, 2026. Custom MCP servers depend on your ChatGPT plan and workspace settings.
Get your footage into Valmera
| Route | How |
|---|---|
| Studio upload | Upload in Valmera Studio, then ask ChatGPT to open that project. Files attached to a ChatGPT message don't reach Valmera. |
| A link | Paste a link and ChatGPT imports the clip, song or image with fetch_url. |
| Codex | Codex has a shell, so it can upload a local file itself with upload_start and upload_finish. |
Main videos can be MP4, MOV, M4V, MKV or WebM, up to 100 GB and 3 hours. Valmera indexes each one once (word-level transcript with speaker labels, shots, silences and on-screen text), and every edit reuses that index.
Your first edit
Describe the outcome; ChatGPT picks the tools. It can ask you to approve actions that change your project; approve them to let it edit.
List my Valmera projects and their IDs.
Open my latest project. Remove the dead air and filler words, add karaoke-style captions, make it vertical for Reels, render a preview and list every change you made.
Find the three strongest 30–60 second moments in this interview and turn each one into a captioned vertical short.
Export the version I approved as a full-quality MP4 and give me the download link.
Review and export the MP4
ChatGPT renders a preview with render_preview, follows the job with wait_for_job, and can return the preview link or sampled frames for review. Revise in plain English; every change writes a new version of a non-destructive edit decision list, so any cut can be restored.
When you approve it, ask for the final video. ChatGPT calls export_final for that project and version, waits for the job, then gets the download link with download_url. The export is a full-resolution H.264 MP4 rendered from your original file, ending with a brief Valmera end card. You can also export from Studio.
Refresh the tools after an update
ChatGPT caches a connection's tool list. When Valmera adds tools, open the connection on chatgpt.com/plugins, select Refresh, and start a new conversation.
Codex
In the Codex CLI, add the server and sign in:
codex mcp add valmera --url https://valmera.io/mcp/server
codex mcp login valmeraIn the Codex IDE extension, open the gear menu → MCP servers → Add server, choose Streamable HTTP, paste the URL, restart the extension and select Authenticate. More clients are on the setup page.
Troubleshooting
| Problem | Fix |
|---|---|
| No option to create a custom MCP server | Turn on Developer mode in Settings → Security and login, or ask your workspace admin to allow custom MCP servers. |
| ChatGPT can't connect | Check the URL is exactly https://valmera.io/mcp/server and authentication is OAuth. Delete the connection and create it again. |
| “Connecting apps to Valmera is not enabled for this account” | Sign in with the Valmera account that has your subscription. MCP access is included with eligible subscriptions. |
| A render says STILL RUNNING | Normal for slow work. ChatGPT calls wait_for_job until the job finishes; it shouldn't start the render again. |
| Editing is refused because the project is busy | The project is open for editing in Studio. Let that edit finish; Studio and MCP can't edit one project at the same time. |
Limits
- Valmera edits footage you provide. It doesn't generate video from a prompt.
- Captions are burned in, with no SRT or VTT file. No direct publishing to YouTube or TikTok.
- Over MCP, ChatGPT edits with the tools directly; it can't hand the job to Valmera's in-house agent.
The full list is on the MCP server page.
Frequently Asked Questions
Sources
ChatGPT and Codex steps were checked against OpenAI's documentation on October 6, 2026.
Made with Valmera: real exports
Connect ChatGPT to Valmera
Creating an account and uploading are free. Editing requires a subscription; MCP access is included with eligible plans.
Connect your assistant →