Free Slicer MCP server
Slicer MCP server: your AI assistant edits for free
Connect Slicer to any client that supports MCP and ask: “make a Reel from interview.mp4”. The assistant reviews the frames and writes the plan, and Slicer checks it with its director’s rules, suggests what to fix and builds the video — for free. Only Slicer’s AI features are paid, and only after the assistant tells you the price.
- First project free
- No card needed
- Ready in minutes
New — your own director
Your own director: your assistant writes the plan, Slicer checks it
Your assistant reviews the frames and writes the plan — Slicer checks it with its own director’s rules on the server and suggests what to fix. The plan becomes a video in your project right away. Your AI does the thinking on your own subscription — so the edit is free.
- Your assistant gets your materialSpeech word by word, pauses, marked retakes, your notes and the plan schema — everything it needs to write the plan.
- The assistant reviews the framesFrames from your footage arrive as images right in the tool response, up to 16 at a time. The assistant sends what it saw along with the plan.
- Director rules stay on Slicer’s serverSlicer checks the plan when it arrives: it fixes word boundaries itself and briefly suggests the rest. The assistant fixes the plan and resubmits.
- Free, on your subscriptionThe material, frames, plan checks and the build cost zero tokens. Speech recognition is at cost, and the assistant names the price before it runs.
The workflow is in the direct_reel prompt. Slicer’s director rules are never sent to the assistant: they run only on our server.
Your AI does the editing, Slicer gives it the editor
Editing costs zero tokens
An Edit it myself project, your own director, edits, caption style, export and publishing over MCP are free.
The AI you already have
Your assistant directs the video by your instructions, using Slicer editor commands, and our director’s rules check the plan.
The whole editor as commands
Cuts, order, hook, callouts, push-ins, emoji, transitions, color, music, voice and inserts — just like in the app.
Slicer AI features on request
AI editing by the Slicer director, captions from speech, the built-in AI assistant and translation — only when you ask.
Price first, then the run
A paid feature returns its price first. The assistant tells you and runs it only after your “yes”.
One address, sign-in in the browser
https://api.slicer.media/mcp over Streamable HTTP, sign-in with OAuth 2.1 or an API token. Nothing to install.
How it works
How to start editing over MCP
- 1
Add the server
In your client’s MCP settings, add https://api.slicer.media/mcp. Steps for popular clients are in the docs.
- 2
Grant access
Sign in to Slicer in the browser and approve access. The monthly spend limit applies only to paid AI features.
- 3
Ask for a video
For example: “edit a Reel from these files”. The assistant uploads the files, reviews the frames, writes the plan, applies Slicer’s suggestions, exports the video and sends you a link to it.
Documentation
The MCP server in the docs
Step-by-step connection, OAuth or token sign-in, your own director, every tool with its parameters, ready-made conversation scenarios, and what is free and what is paid.
- Your own directorMaterial, frames, the plan and suggestions
- ConnectingServer address and client setup
- ToolsWhat the assistant can do in Slicer
- Sign-in: OAuth and tokenConsent window, access and limit
- Example scenariosFrom a file to a published video
- Security and costsPrice confirmation, limits and scopes
How your assistant edits in the Slicer editor
MCP (Model Context Protocol) is an open protocol AI assistants use to connect external services. Through the Slicer MCP server your assistant works in the same editor you use in the app. It creates an Edit it myself project — the media lands on the timeline in order, with no automatic edit — reads the video as text and edits it with commands.
Your AI makes the decisions and Slicer carries them out: it joins the fragments, draws graphics and captions and renders the final file. That is why editing over MCP is free — you edit with the AI you already have. Want a whole video from a plan? The assistant works as your own director. Want precise tweaks? It edits with commands. For questions about Slicer it reads the documentation: it can search and read its pages.
What your assistant does for free
All of these are MCP tools and editor commands — the same ones as manual edits in the app:
- your own director: the material and source frames, plan checks by the Slicer director’s rules and the build
- an Edit it myself project from video, photos and audio — from your computer or a link
- cutting and reordering fragments, removing silent pauses — no speech recognition needed
- a hook, callouts, titles, emoji and camera push-ins
- transitions, color and auto balance, music, voice and inserts over the video
- caption style: template, font, accent and position
- export with file settings and publishing to connected social accounts
Only Slicer’s AI features are paid — and only with your consent
Some jobs Slicer handles with its own neural networks. The assistant offers them but never starts them on its own: without a confirmed amount a paid tool only returns the price. The assistant tells you the price and runs the feature with a second call — after your “yes”.
A monthly spend limit applies on top: 200 tokens by default, set by you in the consent window. The paid features are:
- AI editing by the Slicer director — if you ask for it: a finished video the director checks itself
- captions from speech — at cost per minute of sound
- Slicer’s built-in AI assistant — one-command edits, at cost
- caption translation into another language — at cost
Your files and your video stay yours
The assistant sends footage to a one-time upload link or imports it from a public https address. It shares the finished video as a link to the project in Slicer: there are no public download links without sign-in.
Frames for your own director arrive as images right in the tool response and go only to the AI you connected. No file links are created along the way.
Deleting and stopping a project are marked as destructive — clients that respect these hints ask before calling them.
Pricing
Pay for reels, not a subscription
The in-app currency is tokens. A reel up to 90 seconds costs 24 tokens: it includes a finished reel checked by the director, editor tweaks and export.
Free start
0 ₽
- 1 project
- A full AI edit, checked by the director
- “Slicer” watermark
- Buy any pack — we re-render it without the watermark for free
Edits, color, on-device rendering and export are free. AI operations in the editor are at cost, with the price shown before you start. If an edit fails, tokens come back automatically.
Questions and answers
How much does editing over MCP cost?
Editing is free: an Edit it myself project, your own director, edits, style, export and publishing cost zero tokens. Only Slicer’s AI features are paid: AI editing by the director is 24 tokens per video, and your first video is free; captions from speech, the built-in AI assistant and translation are at cost. The assistant names the price before anything runs.
How is your own director different from Slicer AI editing?
Slicer AI editing is done by our director: it is a paid option for people without their own AI assistant: the director builds the video and checks the finished result itself for 24 tokens, and your first video is free. With your own director, your assistant writes the plan, and Slicer checks it with the same director’s rules and suggests what to fix. Your AI thinks on your own subscription, so the edit is free.
Does the assistant see Slicer’s director rules?
No. Slicer’s director rules run only on our server. The assistant gets your material — speech word by word, pauses, retakes and your notes — and the plan schema, and reviews the frames itself. When it submits the plan, Slicer fixes details like word boundaries and returns short suggestions, such as “two takes of one thought — keep the best one”. The assistant fixes the plan and submits it again.
Are captions free too?
Changing the caption style is free. Captions that follow the speech appear after Captions from speech — it recognizes the speech and is charged at cost per minute of sound. The assistant adds a hook, callouts and titles with your own text for free.
Can the assistant spend tokens without asking?
No. Without a confirmed price a paid feature only names the price. On top of that there is a monthly spend limit, and without the Run montage access the assistant spends nothing.
Which AI assistants work with it?
Any client that supports MCP over Streamable HTTP — chat assistants, code editors and AI agents. Step-by-step setup for popular clients is in the docs.
Do I need an API token?
Not necessarily. Usually sign-in goes through OAuth: the client opens a browser, you sign in to Slicer and approve access. If a client can’t do OAuth, give it an API token from Profile → For developers.
Where do I see the finished video?
The assistant sends a link to the project in the web app: you can watch the video there, refine it in the editor and download it.
Your first reel is free
Sign up in a minute. No card needed.
- First project free
- No card needed
- Ready in minutes