Summarize a clip using Tagstore
GET/q/:hash/rep/summarize_v2
Generates a structured summary for a time-bounded clip of a content
object by fetching tags from the Tagstore service. Unlike
/rep/summarize, this endpoint supports multiple quality levels
that select different LLM backends and prompt strategies, enabling a
trade-off between latency and output richness.
Quality levels:
| Level | LLM backend | Use case |
|---|---|---|
test | Lightweight testing model | Fast iteration during development |
default | Lightweight Ollama model | Production general-purpose summaries |
high | Qwen3 thinking model (vLLM) | High-quality, reasoning-heavy summaries |
The prompt is automatically adapted based on the content title metadata
and detected persons (cast references) associated with the hash.
A configurable padding (in milliseconds) is applied symmetrically
around start_time and end_time before querying Tagstore, so
context at clip boundaries is not lost.
Request
Responses
- 200
- 400
- 401
- 404
- 422
- 500
Clip summary generated successfully.
Missing or invalid request parameters.
The provided authorization token is missing, malformed, or has expired. Ensure the token is a valid Content Fabric bearer token supplied as the authorization query parameter.
No tags were found for the requested content or time range.
Tags could not be retrieved from Tagstore (e.g. network error or Tagstore returned an error).
Internal server error (e.g. LLM unavailable).