CloneViral MCP Server

The official Model Context Protocol server for CloneViral. It gives an agent 22 tools: AI image and video generation, Director Studio, video analysis and dubbing, and Viral Flow graphs.

Prefer the terminal? The same API is a CLI: @cloneviral/cli installs cviral, which can also upload a local file and download a flow's finished assets — things this server cannot do for you.

Which one do I want?

| | | |---|---| | Claude Code / Codex user | Install the plugin. No npm install, and it picks up new tools automatically. | | Any other MCP client | This package, over stdio. See Setup below. | | Scripts, CI, your own shell | @cloneviral/cli |

/plugin marketplace add https://cdn.cloneviral.ai/plugins/claude-marketplace.json

Setup

1. Generate an API key at www.cloneviral.ai/mcp-skills
2. Add to your Claude Desktop config (~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "cloneviral": {
      "command": "npx",
      "args": ["-y", "@cloneviral/mcp-server"],
      "env": {
        "CLONEVIRAL_API_KEY": "cv_your_api_key_here",
        "CLONEVIRAL_API_URL": "https://www.cloneviral.ai"
      }
    }
  }
}

3. Restart Claude Desktop

Available Tools

| Tool | Description | |------|-------------| | Viral Flow | | | cloneviral_list_flows | List your saved Viral Flow graphs | | cloneviral_get_flow | Inspect a flow: nodes in execution order, what can run, estimated credit cost | | cloneviral_create_flow | Create an empty Viral Flow graph | | cloneviral_flow_add_node | Add a node (brief, script/image/video/audio input, media, image, image-edit, video, voice, music, merge, output) | | cloneviral_flow_connect | Connect one node's output to another's input, matched by data type | | cloneviral_flow_run_node | Run one node — image and video return immediately, then poll | | cloneviral_flow_node_status | Check a running node and get its output URL once finished | | cloneviral_flow_list_assets | List every asset in a flow, uploaded and generated | | cloneviral_flow_upload_asset | Import media from a public HTTPS URL (local files: use the cviral CLI) | | Generation | | | cloneviral_list_videos | List your recent generated videos | | cloneviral_get_credits | Check credit balance and subscription | | cloneviral_generate_image | Generate or edit AI images (Nano Banana 2 / Pro, Seedream 5, FLUX 2 Pro, GPT Image 2 / 2.5, Midjourney v8.2) | | cloneviral_generate_video | Generate AI videos — text-to-video, image-to-video, reference-to-video | | cloneviral_check_generation_status | Poll video or image generation status by prediction ID | | cloneviral_list_cinema_presets | List Director Studio looks, camera moves and shot sizes | | cloneviral_generate_cinema_frame | Generate a cinematic still frame from a look + shot size | | cloneviral_animate_cinema_shot | Animate a frame with a camera move (crash zoom, orbit, dolly zoom, drone…) | | cloneviral_analyze_video | Analyze a source video to extract style & structure | | cloneviral_build_remix_variants | Build remix variant prompts from a video analysis | | cloneviral_translate_video | Translate and AI-dub a video into another language | | cloneviral_check_translate_status | Poll a translation job by job ID | | cloneviral_get_characters | List your saved AI characters |

Image Models

Pass a slug as model. The /edit variants take an image_url to modify.

| Family | Slugs | |--------|-------| | Nano Banana | fal-ai/nano-banana-2, fal-ai/nano-banana-2/edit, google/nano-banana-pro | | Seedream 5 | fal-ai/bytedance/seedream/v5/lite/text-to-image, .../lite/edit, .../pro/text-to-image, .../pro/edit | | FLUX | black-forest-labs/flux-2-pro | | GPT Image | openai/gpt-image-2, openai/gpt-image-2/edit | | GPT Image 2.5 | openai/gpt-image-2.5-flare/text-to-image, .../flare/edit, openai/gpt-image-2.5-sunburst/text-to-image, .../sunburst/edit | | Midjourney | midjourney/v8.2, midjourney/v8.2/edit |

Video Models

Pass a slug as model. Use -i2v for image-to-video and -omni / -ref for reference-to-video.

| Family | Slugs | Duration | Resolution | |--------|-------|----------|------------| | Veo 3.1 | veo3.1, veo3.1-fast | 3–20s | sd hd fhd uhd | | Kling 3.0 | kling3-standard, kling3-pro, kling3-4k (+ -i2v) | 3–20s | sd hd fhd uhd | | Seedance 2.0 | seedance2, seedance2-i2v, seedance2-omni (+ -fast, -mini) | 3–20s | sd hd fhd uhd | | Seedance 2.5 | seedance2.5, seedance2.5-i2v, seedance2.5-omni | 4–30s single pass | sd hd | | FLUX 3 | flux3, flux3-i2v, flux3-keyframes, flux3-extend | 5–20s | hd (720p) fhd (1080p) | | MiniMax H3 | minimax-h3, -i2v, -ref | 5–15s (whole seconds) | 2k only | | MiniMax H3 Max | minimax-h3-max, -i2v, -ref | 5–15s | up to 768p |

Seedance 2.5 also accepts 21:9, and seedance2.5-omni takes up to 50 mixed reference_images / reference_videos / reference_audios. seedance2.5-i2v always matches its source image's aspect ratio. minimax-h3-ref takes up to 9 images, 3 videos and 3 audios. FLUX 3 has native audio and uses auto as its aspect-ratio sentinel (not Seedance's adaptive); flux3-keyframes pins up to 10 images to exact frame positions, and flux3-extend continues a clip under 15s / 50 MB. H3 Max is capped at 768p but costs about half of H3 per second. End frames (end_image_url) work on veo3.1*, flux3-i2v, seedance2.5-i2v and minimax-h3-i2v.

Viral Flow

Viral Flow is a node graph: each node is one generation step, wired output to input.

1. cloneviral_create_flow — an empty graph, or cloneviral_list_flows to pick one
2. cloneviral_flow_add_node — add each step (image, video, voice, music, merge, output)
3. cloneviral_flow_connect — wire them together
4. cloneviral_get_flow — execution order, what is runnable, and the credit estimate
5. cloneviral_flow_run_node — run a node, then cloneviral_flow_node_status until done
6. cloneviral_flow_list_assets — collect the output URLs

Run nodes in dependency order; cloneviral_get_flow tells you which are runnable. cloneviral_flow_upload_asset takes a public HTTPS URL — for a file on disk, use cviral flow upload from the CLI package.

Director Studio Workflow

Director Studio is preset-driven: a look grades the still frame, a camera move animates it. The preset wording is composed server-side, so a shot directed from the MCP server or the CLI comes out identical to one directed in the web studio.

1. cloneviral_list_cinema_presets — 15 looks, 24 camera moves, 5 shot sizes
2. cloneviral_generate_cinema_frame — scene + look + shot size → a graded still
3. cloneviral_animate_cinema_shot — that frame URL + camera move → the shot

Step 2 is optional: pass any image URL to step 3 to animate a frame you already have. Each move carries a recommended duration, used automatically unless duration is set.

Remix Workflow

To clone and remix a viral video, use these tools in sequence:

1. cloneviral_analyze_video — Pass the TikTok/YouTube URL to analyze its style
2. cloneviral_build_remix_variants — Generate N variant prompts from the analysis
3. cloneviral_generate_video — Generate each variant as a new video
4. cloneviral_check_video_status — Poll until generation is complete

Video Translation Workflow

1. cloneviral_translate_video — Submit source URL + target language → returns a job ID
2. Check job status at GET /api/video-translate/jobs/{jobId}

Development

npm install
npm run dev    # Run with ts-node
npm run build  # Compile to dist/
npm start      # Run compiled server

src/ is the source of both this package and @cloneviral/cli; the CLI package copies the two files it needs out of dist/ at pack time.

License

MIT