Loading ToolBench
The FIRST MCP server for Kling AI video generation! Generate, extend, lip-sync AI videos, virtual try-on, and manage your account directly from Claude.
Server has 11 tools with complete input schemas and descriptions. Naming follows verb_noun convention (generate_video, check_video_status, etc.). Descriptions are present and moderately detailed (avg ~150 chars), explaining what each tool does. However, several tools lack output schema documentation, parameter descriptions are sometimes generic, and error handling guidance is minimal. No tool annotations (readOnlyHint/destructiveHint) despite clear risk classifications. Some parameters like 'model_name' repeat across tools without clear differentiation of supported versions per tool.
Check the status of a video generation task
Create a lip-sync video by syncing audio or text-to-speech with a video. Supports both audio file input and text-to-speech generation with customizable voice settings.
Apply visual effects to images to create short animated videos with various effect scenes like hug, kiss, heart gesture, squish, expansion, fuzzyfuzzy, bloombloom, and dizzydizzy.
Download a generated video to your local system
Extend a video by 4-5 seconds using Kling AI. This feature allows you to continue a video beyond its original ending, generating new content that seamlessly follows from the last frame. Perfect for creating longer sequences or adding additional scenes to existing videos.
Generate images from text prompts using Kling AI image generation model
No output schemas documented for any tool. LLMs cannot predict response structure, forcing them to guess at field names and types for downstream chaining.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint) despite clear risk classifications (WRITE vs READ_ONLY). Agents cannot distinguish safe from destructive operations.
Minimal error handling guidance. Tools return raw API responses without recovery hints. E.g., 'task_id not found' should suggest 'Call list_tasks() to find valid task IDs.'
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 57 | 2026-07-28+ | v2 |
Generate a video from a reference image using Kling AI. Pass a public URL or local file path as image_url — local files are automatically base64-encoded. Image must be JPG/JPEG/PNG, max 10MB, min 300px. At least one of image_url or image_tail_url is required.
Generate a video from text prompt using Kling AI
Get your Kling AI account balance and resource package information
List all video generation tasks with optional filtering by status, date range, and pagination
Generate virtual try-on images showing how clothing looks on a person using Kling AI's virtual try-on model
Parameter descriptions lack constraint details. E.g., 'prompt' says 'max 2500 characters' but no validation guidance. 'duration' enum is string not number, inconsistent with 'cfg_scale' (number).
Repeated 'model_name' enum across tools (9 versions) without per-tool documentation of which models support which features (e.g., cfg_scale not supported on v2.x).