Agent-native video composition API: compose footage into production-quality video over MCP.
Full LLM thinking from the 4-phase benchmark pipeline.
{
"service_type": "platform",
"base_url": "https://cueframe.ai",
"auth_method": "oauth2",
"auth_config": {
"description": "OAuth 2.0 with PKCE for MCP and REST; also supports API keys and a keyless pay-per-call (x402) model for agents",
"issuer": "https://api.cueframe.ai",
"pkce": true,
"scopes": "unspecified; see OAuth metadata endpoint",
"discovery": [
"https://api.cueframe.ai/.well-known/oauth-authorization-server",
"https://api.cueframe.ai/.well-known/oauth-protected-resource"
],
"alternatives": [
{
"type": "api_key",
"description": "REST authentication via API key (see API quickstart)"
},
{
"type": "x402_pay_per_call",
"description": "Keyless, pay-per-call access for AI agents; no account or API key required"
}
]
},
"endpoints": [
{
"url": "https://api.cueframe.ai/v1/openapi.json",
"description": "Machine-readable OpenAPI specification for the REST API"
},
{
"url": "https://api.cueframe.ai/v1/mcp/x402",
"description": "Public hosted MCP server (Streamable HTTP). Free initialize and tools/list; payment required for data operations"
},
{
"url": "https://api.cueframe.ai/.well-known/oauth-authorization-server",
"description": "OAuth issuer and PKCE endpoints"
},
{
"url": "https://api.cueframe.ai/.well-known/oauth-protected-resource",
"description": "Hosted MCP OAuth discovery"
},
{
"url": "https://docs.cueframe.ai/docs/api",
"description": "API reference covering media, compositions, rendering, sessions, brand kits, and webhooks"
},
{
"url": "https://docs.cueframe.ai/docs",
"description": "Documentation for MCP, CLI, and REST usage"
},
{
"url": "https://cueframe.ai/api.md",
"description": "REST quickstart in Markdown: auth, example, async jobs, safe retries"
},
{
"url": "https://cueframe.ai/cli.md",
"description": "CLI quickstart: auth, rendering saved projects, batch scripts, errors"
}
],
"pricing_model": {
"type": "freemium",
"details": {
"free_tier": "Free desktop editor for macOS and Windows",
"cloud": "Usage-based cloud credits for rendering",
"pay_per_call": "Keyless pay-per-call (x402) model available for agents, no account required",
"notes": "Pricing page at https://cueframe.ai/pricing"
}
},
"rate_limits": {
"documented": true,
"notes": "Rate limits are described in the API policies page at https://cueframe.ai/developers/policies; specific numeric limits not enumerated in the provided content"
},
"capabilities": [
"Deterministic video composition and rendering (same input produces same output)",
"Surgical, incremental edits via structured edit documents (JSON)",
"Author custom React components with GSAP and Three.js instead of fixed template galleries",
"MCP server for agent-native video editing (discover, author, validate, render)",
"REST API for programmatic video composition and rendering",
"CLI for terminal-based and batch video workflows",
"Free local desktop video editor for macOS and Windows",
"Web app at app.cueframe.ai",
"Template library with editable project files",
"Asynchronous rendering jobs with safe retries",
"Brand kits for consistent styling",
"Webhooks for render notifications",
"Media management endpoints",
"Session-based composition authoring",
"Free validation of edits before rendering",
"AI-assisted features: captions (transcription, timing, styling, content-aware placement), face-aware smart crop, headless rendering for agents/CI",
"Pay-per-call x402 access for agents with no account or API key"
],
"raw_analysis": "CueFrame (https://cueframe.ai) is an agent-native, deterministic video composition and rendering platform. Unlike AI video generation models that synthesize pixels from prompts, CueFrame takes a structured edit expressed as data (JSON, often paired with custom React components using GSAP and Three.js) and renders production-quality video consistently—same input yields the same output. Edits are surgical: changing one field changes only that element. The platform exposes multiple interfaces: a REST API (OpenAPI spec at https://api.cueframe.ai/v1/openapi.json), a hosted MCP server for agent integration (https://api.cueframe.ai/v1/mcp/x402), a CLI, a free desktop editor for macOS/Windows, and a web app (app.cueframe.ai). Authentication is OAuth 2.0 with PKCE for MCP and REST, with API keys for REST and a keyless x402 pay-per-call model for agents. The product targets developers and AI agents building automated video pipelines (e.g., social media crops, captioned marketing videos, dynamic content generation). Maturity appears moderate-to-high: it has OpenAPI, OAuth discovery, published policies, changelog, guides, templates, and an active launch (guides include a how-we-rend3/3 tests passed
| Test | Endpoint | Status | Latency |
|---|---|---|---|
| website_uptime | GET / | 200 | 212ms |
| robots_txt | GET /robots.txt | 200 | 178ms |
| llms_txt | GET /llms.txt | 200 | 45ms |
{
"overall": 82,
"dimensions": {
"token_efficiency": 9.0,
"first_try_success": 9.0,
"response_parseability": 9.5,
"error_clarity": 8.0,
"doc_quality": 8.5,
"auth_simplicity": 10.0,
"latency": 10.0,
"consistency": 8.5
},
"pricing_normalized": {
"model": "freemium",
"free_tier": "Free local desktop editor (macOS/Windows)",
"cloud": "Usage-based credits",
"agent_friendly": "Keyless x402 pay-per-call with no account required",
"clarity": "Pricing page available at /pricing; specific per-render and credit costs not enumerated in provided data"
},
"issues": [
"Exact credit/render pricing not enumerated in the provided data, requiring agents to fetch /pricing for concrete numbers",
"No explicit status page referenced, so uptime/consistency perception relies on observed 200s rather than published SLA",
"Cloud rendering still implies usage-based credits, which adds a cost-tracking burden for autonomous agents unless the x402 path is used"
],
"recommendations": [
"Highlight the x402 keyless pay-per-call path as the default agent onboarding route — it removes the account/API-key step entirely",
"Expose machine-readable pricing (per-render, per-minute, per-credit) in llms.txt or a JSON endpoint so agents can budget without scraping",
"Publish a status/uptime endpoint or page to strengthen consistency signals for agent orchestration",
"Surface MCP server connection instructions prominently in llms.txt for one-shot agent setup"
]
}Show your live agent-readiness score on your own site. Free, no auth — it updates as your score changes.
<a href="https://prowl.world/service/cueframe">
<img src="https://prowl.world/badge/cueframe.svg" height="56" alt="Agent-readiness on Prowl">
</a>
See operational metrics, LLM evaluations, agent readiness, and more.
Open in Dashboard