SkillAgentSearch skills...

kie-mcp

MCP server for kie.ai — 45+ image, 70+ video, 20+ audio models with built-in Averiguare model intelligence. Dual-mode (stdio + HTTP) for Claude Code & Cowork.

Install / Use

claude mcp add elibarnett -- npx -y github:elibarnett/kie-mcp

If the server publishes to npm under a different name, use that package instead — check the repo README.

About this skill
🔌

MCP Server

Model Context Protocol server

Quality Score

84/100

Supported Platforms

Claude Code
Claude Desktop

Our assessment of kie-mcp

kie-mcp scores 84/100 on our quality scale, 616th of 947 AI & Machine Learning skills we index.

Its MCP Server is 17 KB long, well organised into 24 sections with 11 code examples: a thorough specification that gives an agent plenty to work with.

It has 10 GitHub stars, so there is little community track record yet; judge it on its content.

Substance
30/30
Structure
20/20
Description
15/15
Adoption
4/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated today, so kie-mcp is actively maintained.
  • Our last check on 2026-11-10 found the source still online.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 97/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-11-10. Automated pattern scan on 2026-11-09. It catches known dangerous patterns, not every risk — read a skill before letting an agent act on it.

kie-mcp compared with similar skills

All 4 of these similar skills score higher than kie-mcp; compare them before choosing.

SkillScoreStarsUpdatedFormat
kie-mcp (this skill)by elibarnett8410todayMCP Server
claude-memby thedotmack100117.6ktodayCLAUDE.md
taste-skillby Leonxlnx100116.8k8d agoCLAUDE.md
Agent-Reachby Panniantong100112.3k8d agoCLAUDE.md
Understand-Anythingby Egonex-AI10092.8k1d agoCLAUDE.md

Frequently asked questions

How do I install kie-mcp?
Run claude mcp add elibarnett -- npx -y github:elibarnett/kie-mcp. The install tabs above show the steps for each supported agent.
Which AI agents does kie-mcp work with?
It is written for Claude Code and Claude Desktop, as a MCP Server file. Other agents that read the same format can often use it too.
Is kie-mcp safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 97/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is kie-mcp still maintained?
The repository was last updated today, so kie-mcp is actively maintained.

kie-mcp

A comprehensive Model Context Protocol server for the kie.ai generation API. Gives Claude (and any MCP client) access to 60+ image models, 85+ video models, and 20+ audio tools with deep model intelligence built in.

Why this exists

Most MCPs are thin API wrappers. This one is different:

  • Vertical profiles (NEW in 5.0) — per-domain playbooks: profile_brief returns the intake questions a professional would ask, model routing per deliverable with live costs, per-model prompt formulas, and multi-tool workflows. Profiles double as MCP prompts (/kie-art:architecture in Claude Code). Verticals (10): architecture & interiors, video game assets, advertising & marketing, web & software product imagery, film & storyboarding, product photography & e-commerce, brand & graphic design, editorial & publishing, short-form social video, and audio branding & music.

  • Deep research embedded — Every major model has a research field with verdicts, prompt techniques, weaknesses, cost-efficiency analysis, and competitor comparisons. Researched by Averiguare, our model intelligence agent.

  • Cost-aware — Every model has pricing in credits and USD. The MCP tells you the cheapest option for your use case.

  • Smart filtering — list_models filter="lip sync" or filter="architecture" or filter="cheapest video" — searches across capability tags, descriptions, AND research fields.

  • Dual-mode transport — stdio for local Claude Code, HTTP Streamable for remote Cowork/cloud usage.

What you can do with it

Just ask Claude things like:

  • "Generate a brand presentation board for a perfume launch" — picks GPT Image 2 (best for text-heavy layouts)
  • "Make a 10s video of fruit scarecrows defending against crows, Pixar style" — recommends Veo 3.1 or Wan 2.7
  • "Generate music for a fantasy adventure game" — Suno V5
  • "Lip-sync this audio to my character image" — Kling AI Avatar or Infinitalk
  • "Upscale this video to 4K" — Veo 4K upscale or Topaz
  • "Replace the wall color in this room photo" — Flux Kontext Pro (best for surgical edits)

Model coverage

Image (60+)

  • OpenAI: GPT Image 2.6 Flare / Sunburst / Economy (NEW Nov 2026; Economy 5 cr flat, animated GIF), GPT Image 2.5 Flare / Sunburst (NEW — 1K-4K, transparent background), GPT Image 2, GPT-4o Image, GPT Image 1.5
  • Google: Nano Banana 2.1 (NEW, 4 cr), Nano Banana 2 / 2 Lite / Pro / Edit / Original, Imagen 4 (Fast/Standard/Ultra)
  • Black Forest Labs: FLUX 3 Klein T2I + Edit (NEW, 1.6 cr), Flux Kontext Pro/Max, Flux 2 Pro/Flex
  • ByteDance: Seedream 3.0 / 4.0 / 4.5 / 5.0 Lite
  • Alibaba: Wan 2.7 Image / Image Pro
  • Ideogram: v3, Character, Edit, Remix, Reframe
  • xAI Grok Imagine Image 2.0 (#2 Arena T2I + edit; free segment map → region-targeted edit chain; whole-image edits of ANY uploaded image)
  • ByteDance Seedream 5.0 Pro (NEW — T2I/I2I + layer decomposition: split any image into layer files)
  • Qwen Image 3.0 / 3.0 Pro (NEW — seed, negative prompts, 2K at the 1K price on standard)
  • Qwen Image 2.1 (NEW — transparent background, mask inpainting, 10-ref compositing)
  • Others: Qwen/Qwen2, Z-Image, Grok Imagine 1.x, Recraft, Topaz

Video (85+)

  • Google Veo 3.1: Quality / Fast / Lite (T2V + I2V), Extend, 1080p/4K upscale
  • Alibaba HappyHorse: 1.1 (NEW — T2V/I2V/R2V with native audio + 7-language lip-sync), 1.0 (T2V/I2V/R2V/Video Edit)
  • ByteDance Seedance: 2.5 (30s single takes, 1080p, draft→1080p upgrade) / 2.0 / 2.0 Fast / 2.0 Mini (1.5 Pro retired upstream Oct 2026)
  • Kuaishou Kling: 4.0 Flash (NEW — 720p, 17 cr/s), 4.0 (NEW — reference-to-video with named @subjects/voices/video refs, keyframe-to-video, 3-30s), 3.0 Omni "O3" (NEW — per-shot multi_prompt scripting, 4K, video Transformation), 3.0, 3.0 Turbo, 2.6, V2.5 Turbo, V2.1 Master/Pro/Standard, AI Avatar
  • Alibaba Wan: 3.0 + 3.0 Prime (NEW — unified prompt-or-media, audio), 2.7 (T2V/I2V/Edit/R2V), 2.6, 2.5, 2.2 Turbo, Animate
  • Google Gemini Omni: Video + 1.1 Flash (NEW — first→last-frame mode, 360p-4K; image/voice/video/character references)
  • MiniMax Hailuo: H3 (NEW — 2K + native stereo audio, image+video+audio references, first→last-frame I2V), 2.3 Pro/Standard, 02 Pro/Standard
  • xAI Grok Imagine: Video 1.5 preview (NEW — I2V with native audio, cheapest audio video), T2V, I2V, Upscale, Extend
  • Avatar / lip-sync: OmniHuman 2.0 (NEW — 14 cr/s, up to 5-minute audio), OmniHuman 1.5 (NEW — audio-driven full-body avatar + free subject-detection + human-identification utilities), Volcengine Video Lip-Sync (NEW — re-dub existing footage), Kling AI Avatar, Infinitalk
  • PixVerse V6 / V6 Max (V6 Max NEW: higher-quality T2V/I2V): T2V, I2V (viral templates), Transition (first→last morph), Fusion R2V (@ref_name), Extend — budget all-rounder with native audio
  • Runway: Aleph, Aleph Edit, Extend
  • Others: ByteDance V1 Pro/Lite, Topaz upscale

Audio (20+)

  • Suno: Music Gen, Extend, Cover, Add Instrumental/Vocals, Replace Section, Lyrics, Sounds, Sound Effects, MIDI, Music Video, Cover Art, Mashup, Persona, Timestamped Lyrics, Boost Style, Vocal Separation, WAV, Custom Voice cloning (experimental)
  • NVIDIA Nemotron 3.5 Transcribe (NEW): speech_to_text default — 0.3 cr per minute, ≤60 min, 32 languages, word timings + speaker labels, translation
  • ElevenLabs: TTS (Turbo 2.5 + Multilingual V2), Text-to-Dialogue V3, Audio Isolation, Speech-to-Text
  • Google Gemini TTS (NEW): style-directed speech, 30 voices, 2-speaker dialogue, inline tone tags — ~4.2 cr/min

Utility

  • File upload (URL or base64)
  • Veo Extend, 1080p Upscale, 4K Upscale
  • Runway Extend
  • Task status, credit check, raw asset listing

Installation

Prerequisites

Setup

git clone https://github.com/YOUR_USERNAME/kie-mcp.git
cd kie-mcp
npm install

Run as stdio MCP (Claude Code, Claude Desktop)

Add to your Claude config (~/.claude.json for Claude Code, or your MCP client's equivalent):

{
  "mcpServers": {
    "kie-art": {
      "command": "node",
      "args": ["/absolute/path/to/kie-mcp/server.mjs"],
      "env": {
        "KIE_API_KEY": "your-kie-ai-api-key",
        "KIE_PROJECT_ROOT": "/optional/path/for/outputs"
      }
    }
  }
}

Or use the Claude Code CLI:

claude mcp add -s user kie-art /usr/bin/env -- KIE_API_KEY=your-key node /path/to/server.mjs

Run as HTTP MCP (Cowork, remote clients)

HTTP mode requires a bearer token and listens on 127.0.0.1 by default.

export KIE_MCP_AUTH_TOKEN=$(openssl rand -hex 32)
KIE_API_KEY=your-key node server.mjs --http --port=3100

Configure your MCP client with the URL and the token:

{
  "mcpServers": {
    "kie-art": {
      "type": "http",
      "url": "http://127.0.0.1:3100/mcp",
      "headers": { "Authorization": "Bearer <your KIE_MCP_AUTH_TOKEN>" }
    }
  }
}

To reach it remotely, put it behind a tunnel (ngrok / Cloudflare Tunnel) or a reverse proxy. The server stays on loopback and the tunnel forwards to it. Add the tunnel's hostname to KIE_MCP_ALLOWED_HOSTS. Treat the token like a password: anyone who has it can spend your kie credits and upload media files from the server's disk.

Security note (5.2.2): earlier versions ran HTTP mode on 0.0.0.0 with no authentication and Access-Control-Allow-Origin: *. Any web page open in your browser, or anyone who had the tunnel URL, could call the tools, including reading local files through upload_file. Upgrade if you use --http.

Environment variables

| Variable | Required | Purpose | |---|---|---| | KIE_API_KEY | yes | Your kie.ai API key | | KIE_PROJECT_ROOT | no | Server-wide default for where generated files are saved (default: server cwd; files go to $KIE_PROJECT_ROOT/kie/assets/raw/). Per-call download_dir (absolute path) on any file-writing tool overrides this | | KIE_MCP_PORT | no | Port for HTTP mode (default: 3100) | | KIE_SUNO_API | no | legacy sends Suno tools to kie's old Suno endpoints. Default since 5.5.0 is kie's new market API (ai-music-api/*), which reports the actual credits charged. MIDI, mashup and voice cloning always use the old endpoints | | KIE_ELEVENLABS_ENABLED | no | 1 re-enables the ElevenLabs tools (paused since 5.4.0 while kie's ElevenLabs integration fails) | | KIE_MAX_CREDITS_PER_CALL | no | Default spending cap per tool call, in credits. Calls whose estimate exceeds it are refused before submission; a per-call max_credits overrides it | | KIE_MCP_AUTH_TOKEN | HTTP mode | Bearer token HTTP clients must send (Authorization: Bearer …). The server refuses to start in HTTP mode without it | | KIE_MCP_HOST | no | HTTP bind address (default 127.0.0.1). Set 0.0.0.0 only inside a container or behind a firewall | | KIE_MCP_ALLOWED_HOSTS | no | Comma-separated extra Host header names to accept (e.g. your tunnel hostname). On loopback, localhost names are always accepted; other hosts are rejected to block DNS rebinding | | KIE_MCP_ALLOWED_ORIGINS | no | Comma-separated browser origins allowed to call the server (CORS). Requests with any other Origin header are rejected. Non-browser MCP clients send no Origin and are unaffected | | KIE_CALLBACK_URL | no | Callback URL sent with Suno generation requests (kie.ai requires the field; results are fetched by polling regardless). Defaults to an inert placeholder — set this only if you want to receive the callbacks yourself | | KIE_MAX_CONCURRENT | no | Max simultaneous task-creation calls (default 4). Excess parallel generations queue inside the server instead of hitting kie.ai's rate limits — parallel tool calls are safe | | KIE_POLL_BUDGET_IMAGE / _VIDEO / _AUDIO / _SPEECH | no | Blocking-mode polling budget per tool category, in seconds (defaults: 600 / 900 / 300 / 300). Per-call max_wait_seconds takes precedence. For long generations prefer wait: false (async mode): the tool returns the task_id immediately; poll with check_task, fetch with download_result |

Tools available

generate_image, generate_video, generate_music, generate_sfx,
generate_tts, generate_gemini_tts, generate_dialogue, generate_sounds, generate_lyrics,
generate_persona, generate_mashup, generate_cover_art,
generate_midi, create_music_video,
prepare_voice_clone, create_voice_clone, regenerate_voice_clone,
create_omni_voice, create_omni_character,
extend_music, cover_audio, upload_extend_audio,
add_instrumental, add_vocals, replace_section,
convert_to_wav, separate_vocals, boost_style,
get_timestamped_lyrics, audio_isolation, speech_to_text,
profile_brief,
list_models, check_task, list_tasks, check_credits, spend_report,
download_result, list_raw_assets, upload_file,
grok_segment_map, grok_image_edit, seedream_layer_decompose,
veo_extend, veo_upscale_1080p, veo_upscale_4k, runway_extend

Cost control

kie bills per call, and a few calls are expensive (1080p/4K video, layer splits, 1080p draft upgrades). Every paid tool takes two optional parameters:

  • estimate_only: true returns the estimated credits from published rates, including resolution/size tiers and video duration, without submitting anything.
  • max_credits: N refuses to submit when the estimate exceeds N. Set KIE_MAX_CREDITS_PER_CALL to apply a default cap to every call.
generate_video { model: "bytedance/seedance-2-5", model_options: { resolution: "1080p", duration: 10 }, estimate_only: true }
→ 💰 ~1580 credits (~$7.90) — 158 cr/s × 10s. Nothing was submitted.

spend_report shows where the credits went. It totals the server's task history over the last N days, by model or by day, using the actual amounts kie charged (it looks up any missing ones), and lists failed tasks separately.

Smart model recommendations

Try these queries in any MC

Truncated for display — read the full file on GitHub.

Related Skills

View on GitHub
GitHub Stars10
CategoryAI
Updated11h ago
Forks6

Languages

JavaScript

Trust signals

97/100

From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.

1 info