Pureframe AI API Documentation

By default, Pureframe AI searches all available signals and returns one relevance-ranked result set. Use modes when your application needs a narrower search.

Set search modes

modes=["video", "transcript", "scene"]

This is the default. Pass a subset to restrict which signals are searched:

Mode	Matches segment type
`video`	`frame`: visual content
`transcript`	`transcript`: transcribed speech
`scene`	`scene`: detected scene labels

# Visual and transcript only, skip scene labels
curl -X POST https://api.pureframe.ai/v1/search \
  -H "Authorization: Bearer pf_..." \
  -F "query=whiteboard diagram" \
  -F "modes=video" \
  -F "modes=transcript" \
  -F "collection_id=col_abc123"

Text visible in a frame, including signs, slides, and captions, can appear as an ocr result. OCR results are not controlled by modes.

How results merge

Pureframe AI combines matches from the selected modes and ranks them by relevance. A segment can rank well because it matches visual content, spoken content, or more than one signal. Do not compare scores from different requests as absolute confidence values. Use the score to order results within a response and validate important matches with the returned timestamp, thumbnail, and text.

Filtering by scene label

Pass labels to restrict results to matching scene labels such as outdoor, whiteboard, or person:

curl -X POST https://api.pureframe.ai/v1/search \
  -H "Authorization: Bearer pf_..." \
  -F "query=demo" \
  -F "labels=whiteboard" \
  -F "collection_id=col_abc123"

Use meta.facets.scene_labels to build a filter interface from the labels represented in the result set.

Precedence when combining image and text

When you send both an image and query, the image provides the primary visual reference and the text narrows the intent. See Image search.

Get Started

Search

Video Processing

Agents

Production

Resources

Combined Queries

Set search modes

How results merge

Filtering by scene label

Precedence when combining image and text

​Set search modes

​How results merge

​Filtering by scene label

​Precedence when combining image and text

Set search modes

How results merge

Filtering by scene label

Precedence when combining image and text