Valmera — agentic AI video editor

Valmera MCP server — the video editing MCP server for real footage

The world's most advanced AI video editor. [Start editing](https://valmera.io/chatgpt-video-editor?utm_source=valmera_mcp_docs&utm_medium=repository&utm_campaign=mcp_chatgpt_workflow#first-brief). Join thousands of creators. Bring your footage, describe the result, and refine the edit through conversation.

Valmera is an agentic AI video editor. This hosted Model Context Protocol server lets Claude, ChatGPT, Gemini CLI, Cursor, VS Code, Codex or any MCP client edit video you already recorded. The assistant reads the transcript and frames, then cuts silences and filler words, adds word-timed captions, reframes to 9:16, adds b-roll, music and sound effects, and makes shorts from long videos. It renders a preview, watches the result, and exports a full-quality MP4 from your original file.

[Connect your assistant](https://valmera.io/mcp/connect?utm_source=valmera_mcp_docs&utm_medium=repository&utm_campaign=mcp_chatgpt_workflow) · [Open an editing brief in Valmera](https://valmera.io/chatgpt-video-editor?utm_source=valmera_mcp_docs&utm_medium=repository&utm_campaign=mcp_chatgpt_workflow#first-brief) · [ChatGPT first-edit walkthrough](docs/CHATGPT.md)

Comparing editors? Use the edit-and-revise evaluation kit: the same source footage, exact prompts, expected source ranges and a blank evidence scorecard.

Start with your own footage and an editing goal. Use the public interactive editing brief builder, then follow the creator workflow briefs for your first project.

Endpoint   https://valmera.io/mcp/server

Transport Streamable HTTP Auth OAuth 2.1 (dynamic client registration + PKCE S256), or a bearer token from https://valmera.io/mcp Registry io.valmera/video-editor

• Product: valmera.io
• Setup for every client: valmera.io/mcp/setup
• Tool reference: valmera.io/mcp/tools
• Live server card: /.well-known/mcp/server-card.json

Connect your assistant

Want to see the output first? Inspect a real edit below, or follow the client-specific connection instructions here.

Claude (claude.ai and Claude Desktop)

Settings → Connectors → Add custom connector → name it Valmera. Paste https://valmera.io/mcp/server and sign in when prompted. Step-by-step guide.

Claude Code

claude mcp add --transport http valmera https://valmera.io/mcp/server

Then run /mcp inside Claude Code and choose Authenticate. Guide.

Cursor — ~/.cursor/mcp.json

{

"mcpServers": { "valmera": { "url": "https://valmera.io/mcp/server" } } }

VS Code (GitHub Copilot agent mode) — .vscode/mcp.json

{

"servers": { "valmera": { "type": "http", "url": "https://valmera.io/mcp/server" } } }

Windsurf — ~/.codeium/windsurf/mcp_config.json

{

"mcpServers": { "valmera": { "serverUrl": "https://valmera.io/mcp/server" } } }

ChatGPT

Add Valmera as a custom MCP server with OAuth at https://valmera.io/mcp/server, complete sign-in, and install the resulting plugin. Select Valmera with @ in a new conversation. Account and workspace permissions apply. Connection and first-edit walkthrough · Guided setup.

Codex CLI

For the connection plus guided setup and three editing workflows, see the ChatGPT and Codex plugin package. The repository supplies a marketplace for supported local clients; public Directory approval is separate.

codex mcp add valmera --url https://valmera.io/mcp/server

codex mcp login valmera

Or add [mcp_servers.valmera] with url = "https://valmera.io/mcp/server" to ~/.codex/config.toml.

Gemini CLI

gemini mcp add --transport http valmera https://valmera.io/mcp/server

Then open Gemini CLI, run /mcp auth valmera and sign in.

See the Gemini CLI editing guide for the current account requirements and a first-edit walkthrough.

Antigravity

Use the native Valmera plugin for the hosted connection and three editing skills. It includes a pinned installation command, a first-edit prompt and an alternative workspace MCP configuration. Antigravity uses serverUrl in .agents/mcp_config.json; do not copy the Gemini CLI configuration unchanged.

Any stdio-only client

{

"mcpServers": { "valmera": { "command": "npx", "args": ["-y", "mcp-remote", "https://valmera.io/mcp/server"] } } }

There are more clients (Zed, Cline, Continue, Goose, LibreChat, MCP Inspector) in docs/CLIENTS.md.

First message to try: *"List my Valmera projects."* An empty list still means the connection works.

Watch a real video edit

This published Valmera shorts example turns selected footage from a 49-minute talk into a 33.24-second vertical story. Compare the same opening before styling and after framing, captions, graphics and color:

Click either image to play its silent 9.2-second excerpt. The first clip already has its source ranges selected; it is not the uncut recording. Watch the complete export with sound to inspect the speech and joins. That historical file lasts approximately 38.27 seconds including its branded ending; its branding does not establish current plan terms.

The complete editing walkthrough includes four intermediate stages, the seven-cut map, the operation record and unresolved review notes. This demonstrates a Valmera shorts workflow, not a timed MCP-client test, a claim that one literal prompt produced every step, or a competitor benchmark.

Footage: The End of the Future with Peter Thiel, Stanford Classical Liberalism Initiative. Neither the speaker nor the source organization is presented as a Valmera customer or endorser.

For a repeatable test on your own shortlist, use the edit-and-revise evaluation kit. For your own recording, start with a creator workflow brief.

Install as a Claude plugin (connector + editing skills)

For a standalone ZIP, connection steps and a first-edit brief, use the Claude plugin guide and version 1.0.2 release. This repository release does not imply approval in Anthropic's public directory.

This repository is also a Claude plugin. It bundles the Valmera connector with three skills that teach Claude a reliable editing workflow: edit-video-with-valmera, turn-long-video-into-shorts and website-demo-video.

In Claude Code:

/plugin marketplace add ABO3SKRALMASAOODI/valmera-mcp

/plugin install valmera@valmera

In claude.ai, go to Customize → Plugins → Add → Add marketplace and enter ABO3SKRALMASAOODI/valmera-mcp. The skills follow the open Agent Skills format, so other agents that read SKILL.md can use them too.

Install a workflow skill in another coding agent

The skills are MIT-licensed Markdown instructions; the hosted editing service requires its own authenticated account and paid editing plan. Installing a skill does not connect MCP or start a paid job.

With Node.js 22.20 or later, install the editing skill into the current project using the pinned Skills CLI:

npx skills@1.7.6 add ABO3SKRALMASAOODI/valmera-mcp@v1.0.1 --skill edit-video-with-valmera -a codex

Use -a claude-code or -a cursor for those clients. Select turn-long-video-into-shorts or website-demo-video instead when that matches the job. Then connect and authenticate Valmera. The skill prompts the assistant to discover current tool schemas, establish the scope and cost expectations, review a preview, and export the approved version.

For an actual result, see the trim-and-reframe example: original motion-graphic footage, selected source ranges, a square crop and the rendered export.

What your assistant can do with it

Use the live server card and your client's current tool discovery for the available tools. Valmera evolves frequently; the examples below explain the workflow rather than fixing a permanent tool count. The reference is in docs/TOOLS.md.

How it works

1. Upload. Upload in Valmera Studio, import a link, or use upload_start and upload_finish from a client that can send files. Check the current upload requirements in Studio and the upload documentation.
2. Index. Valmera transcribes the footage with word-level timestamps and detects shots, silences and on-screen text. index_status reports progress.
3. Edit. Each tool call writes a new version of a non-destructive edit decision list (EDL). The original upload is never modified, and any cut can be restored.
4. Review. render_preview renders a fast proxy preview. watch_video and look_at let the assistant inspect the frames it produced before it repor