# Claude AI Video Generator > Claude can't generate videos on its own — it's a text model with no video model inside. To make video in Claude, add a video tool through a connector (MCP). With the Kenerate connector you paste one URL into Claude, sign in, and ask: Claude writes the prompt, calls `generate_video` on a model like Wan 3.0 or Seedance 2.5, waits for the render with `wait_for_generation` and gives you the clip. Video needs a paid Kenerate plan, the same as in the app. URL: https://kenerateai.com/claude-ai-video-generator Publisher: Kenerate AI (https://kenerateai.com) Last updated: 23 October 2026 ## Key facts - Native video model in Claude: No — text, code and images in, text out - How to add one: Custom connector (MCP) - Works in: claude.ai, Claude Desktop, mobile, Claude Code - Video tools: generate_video, talking_avatar, upscale_video, add_subtitles… - Video models: Wan 3.0, Seedance 2.5 / 2.0, Veo 3.1, Kling 3… - Longest clip: 30 s (Wan 3.0, Seedance 2.5) - Plan needed: Paid plan for video ## Steps 1. Add the connector — In Claude open Customize → Connectors → + Add → Add custom connector. Name it Kenerate and paste https://mcp.kenerateai.com/mcp — or use the “Add to Claude” button on /app/mcp, which fills both in. 2. Sign in to Kenerate — Click Add, then Connect. A Kenerate page opens: sign in and click Allow. No API key needed. Video runs on a paid Kenerate plan. 3. Switch it on in a chat — In a chat click + → Connectors and turn on Kenerate. Claude Code instead: run claude mcp add --transport http kenerate https://mcp.kenerateai.com/mcp, then /mcp → Authenticate. 4. Ask for a video — “Use Kenerate to make a 5-second 9:16 clip of…”. Claude picks the model, runs generate_video, waits for the render and hands you the link; the clip is also saved in your Kenerate history. ## In depth ### Why Claude needs a connector to make video Claude is a language model. It can write a script, a shot list or an animation in code, but it can't render footage. The Model Context Protocol (MCP) is how Claude uses outside tools, and a connector is an MCP server it can call. Add a video connector and the work splits naturally: Claude turns a rough idea into precise prompts — subject, motion, camera, sound — and the video model renders them. Claude also handles the waiting: video jobs run in the background and Claude checks on them until the clip is ready. ### The video tools Claude gets The Kenerate connector has 29 tools. The video ones: - generate_video — text to video, or image to video with a start frame. - talking_avatar — a portrait plus a voice track becomes a speaking video. - motion_control and restyle_video — copy motion from a clip, or restyle footage. - upscale_video, add_subtitles and dub_video — finish and localise a clip. - estimate_cost, pick_model and wait_for_generation — price, choose and track the job. ### The motion prompts behind the start frames above Each still was animated with Wan 3 at 720p for 5 seconds, with sound, on our image-to-video page. Ask Claude for the same: - Painted storm: “Huge waves crash against the cliff, the lighthouse beam sweeps through the storm, the brush strokes shimmer, thunder and surf.” - Product reveal: “The sneaker spins slowly in mid-air and bounces softly on the pink backdrop, playful product reveal, upbeat music.” - Anime countryside: “The little red train chugs across the hills, clouds drift, grass waves in the breeze, whimsical anime style, cheerful music.” - Fantasy art: “The jade dragon slowly uncoils around the temple, mist rolling off its scales, it opens its eyes, deep rumble.” ## FAQ Q: Can Claude AI generate videos? A: Not by itself. Claude reads text, code and images and writes text; it has no video model. Add a video tool as a connector (MCP) and Claude can make clips: it writes the prompt, calls the tool, waits for the render and gives you the result. With Kenerate that tool is generate_video, plus image-to-video, talking avatars and upscaling. Q: How do I add a video generator to Claude? A: Open Customize → Connectors → + Add → Add custom connector, name it Kenerate and paste https://mcp.kenerateai.com/mcp. Click Add, then Connect, sign in to Kenerate and click Allow. Then switch Kenerate on in a chat with + → Connectors. The “Add to Claude” button on kenerateai.com/app/mcp fills in the name and URL for you. Q: Is the Claude AI video generator free? A: Connecting is free — Claude's free plan can add one custom connector and Kenerate doesn't charge for the connection. Making video is not: it needs a paid Kenerate plan, exactly as in the app, and costs credits per second. Wan 3.0 is the cheapest place to start at 15 credits a second at 480p. Images, music and speech follow your plan's normal rules. Q: Does it work in Claude Code and Claude Desktop? A: Yes. A connector added on claude.ai syncs to Claude Desktop and mobile. In Claude Code run claude mcp add --transport http kenerate https://mcp.kenerateai.com/mcp (add --scope user for every project), start claude, type /mcp and pick kenerate → Authenticate. Claude Code can then save the clip links into your project. Q: Which video models can Claude use? A: Through Kenerate: Wan 3.0 (up to 30 s), Seedance 2.5 (up to 30 s, 4K), Seedance 2.0, Veo 3.1, Kling 3 and more of the app's video models. Claude can call list_models or pick_model to choose one for the job, or you can name the model in your message. Q: Can Claude turn my image into a video? A: Yes. Share the image in the chat (Claude uploads it with upload_media) or let Claude make one with generate_image first, then ask for the motion. generate_video uses the picture as the first frame, so products, characters and faces stay as they are. Q: How long are the videos and how long do they take? A: Wan 3.0 and Seedance 2.5 go up to 30 seconds per clip, Seedance 2.0 and Kling 3 up to 15, Veo 3.1 up to 8. Renders usually take a few minutes; Claude checks with wait_for_generation and posts the link when the clip is done. Q: Can Claude add a voice-over, music or subtitles? A: Yes, with separate tools on the same connector: generate_speech for a voice-over, generate_music for a music bed, add_subtitles for captions and dub_video for another language. Several video models also make native sound. Joining clips or mixing an audio track into a video is a local editing step, not an MCP tool. Q: Do I need an API key? A: Not for claude.ai, Claude Desktop or Claude Code — they sign in with OAuth: you click Allow on a Kenerate page. Keys are for apps without sign-in and for scripts. Either way, the connection shows on /app/mcp and deleting it disconnects Claude. Q: What will Claude refuse to make? A: Kenerate applies the same rules as its app whatever the client: content involving minors, sexual content of real people and illegal content are always refused. ## Related pages - Claude AI image generator: https://kenerateai.com/claude-ai-image-generator - Kenerate MCP: https://kenerateai.com/app/mcp - AI video generator: https://kenerateai.com/ai-video-generator - Image to video AI: https://kenerateai.com/image-to-video-ai ## Sources - Kenerate MCP — connect Claude, Cursor, ChatGPT and more: https://kenerateai.com/app/mcp - Claude Help Center — Get started with custom connectors using remote MCP: https://support.claude.com/en/articles/11175166-getting-started-with-custom-connectors-using-remote-mcp - Claude Code docs — connect to MCP servers: https://code.claude.com/docs/en/mcp - Model Context Protocol: https://modelcontextprotocol.io/