---
title: "Re-run AI analysis on an asset"
method: POST
path: "/v1/assets/{id}/analyze"
tags: ["Intelligence"]
---

# Re-run AI analysis on an asset

`POST /v1/assets/{id}/analyze`

## Request

- Base URL: `https://api.picsha.ai`
- URL: `https://api.picsha.ai/v1/assets/{id}/analyze`
- Auth: HTTP bearer

## Path parameters

- `id` string, required

## Request body

- object, nullable
  - `config` object: Processing pipeline options, mirroring ingest config.
    - `auto_tag` boolean: Whether to run AWS Rekognition Vision AI for face and object detection. Defaults to true.
    - `auto_summarize` boolean: Whether to run AI summaries (Claude 3 Haiku for images, Titan/Claude for documents, Amazon Transcribe for audio/video). Defaults to true.
    - `vectorize` boolean: Whether to generate a multimodal embedding vector for similarity search. Defaults to true.
    - `location_lookup` boolean: Whether to reverse-geocode EXIF GPS coordinates via Google Maps API. Defaults to true.
    - `quality_metrics` boolean: Whether to compute image quality metrics during processing.

## Response `202`

Default Response

- object
  - `id` string, required
  - `message` string, required

## Other responses

- `404`: Default Response
- `500`: Default Response

---

[API](https://skmtc.dev/picsha/apis/picsha-ai-api.md) · [All operations](https://skmtc.dev/picsha/apis/picsha-ai-api/llms.txt) · [OpenAPI document](https://skmtc.dev/picsha/apis/picsha-ai-api/revisions/f77c6ac40e70?raw)