Llama 3.2 90B Vision Instruct

Meta's largest vision model, optimized for document understanding and visual reasoning.

⭐ Flagship

Context Window

128K

Input Price

$N/A

per 1M tokens

Output Price

$N/A

per 1M tokens

Family

Llama 3.2

Capabilities

Vision

Function Calling

JSON Mode

Streaming

Reasoning

Image Generation

Input Modalities

textimage

Output Modalities

text

Model Information

Status: active
Access Type: open weights
License: Proprietary
Last Updated: 12/30/2025