Computer vision · GA · v1.0

Vision, Detect, OCR, and Segment in One Call

Computer vision on Simon — detect, OCR, and segment in one call. SAM 3 and specialist slugs for production vision workloads.

DetectionOCRSegmentationCAT-ready

Image in · structured data out · GA

Vision

Detect · OCR · segment

detectocrsegment

invoice

signature

Extracted fields

VendorAcme Supplies

Invoice #INV-2026-0412

Total$4,280.00

Signedyes · 0.94

Regions2 detected

OCR128 tokens

Confidence0.94 avg

POST /v1/chat/completions

"model": "simon-says-vision"

Image in · structured data outone API call

Capabilities

Production vision without glue code

Extract text and objects, run precise segmentation, and drop vision steps into multi-step CAT workflows.

Extract

Extract text and objects from images in one request.

SAM 3

SAM 3 and related slugs for precise masks.

CATs

Vision steps inside multi-step workflows.

Catalog

Browse vision-capable slugs on /models.

Works with

Our lineup of AI models — experts, orchestrators, and direct frontier access.

Compose multi-step workflows — call cat:your-slug in one request.

One OpenAI-compatible API — chat, voice, vision, and our full model lineup.