How it works
When an agent callssearch_videos, results include thumbnail_base64 — a base64-encoded JPEG of the matched frame. Vision-capable models can see this image in the same turn:
- Agent searches for a moment in video
- Pureframe AI returns the matched frame as base64
- Agent reads the visual content and answers follow-up questions
Two ways to connect
MCP — the simplest path for Claude, Cursor, VS Code, Codex, OpenCode, and any other MCP-compatible client. See MCP Overview and Connect an Agent. Direct HTTP / function calling — for OpenAI, Gemini, or any LLM with function calling, fetch the OpenAI-compatible schema and wire it directly:schema directly to client.chat.completions.create(tools=schema) for OpenAI or the equivalent for other providers.
Example: search + vision with Claude
Tool response fields
search_videos returns a list of clips. Each clip:
Every answer an agent gives from these fields is traceable back to an exact clip — see Source-backed Results.