vision

name	vision
description	Call vision models (Doubao, Qwen, OpenAI) to analyze images. Use when you need to understand screenshots, UI layouts, diagrams, or any image content. Supports png/jpg/webp/gif.

vision

Multi-provider vision tool. Call various vision models to describe images. Feed it a prompt + image path, get back a text description.

Quick start

python vision.py [--provider <name>] <image_path> <prompt>

When --provider is omitted, the provider is resolved by: --provider flag > VISION_PROVIDER env > first API key found.

Providers

doubao (豆包 / Volcengine Ark)

• API key: DOUBAO_API_KEY
• Default model: doubao-seed-2-0-pro-260215
• Custom endpoint: DOUBAO_BASE_URL

qwen (通义千问 / DashScope)

• API key: DASHSCOPE_API_KEY
• Default model: qwen-vl-max
• Custom endpoint: DASHSCOPE_BASE_URL
• Available models: qwen-vl-max, qwen-vl-plus, qvq-max

openai (GPT-4o)

• API key: OPENAI_API_KEY
• Default model: gpt-4o
• Custom endpoint: OPENAI_BASE_URL
• Also works with any OpenAI-compatible endpoint.

Configuration

Env Var	Scope	Default
VISION_PROVIDER	Default provider	auto-detect
VISION_MODEL	Override model (all providers)	provider default
{PROVIDER}_MODEL	Override model (per provider)	—
VISION_TEMPERATURE	Response creativity 0–1	0
VISION_MAX_TOKENS	Max response tokens	4096

Examples

# Auto-detect provider from API keys
python vision.py "screenshot.png" "Describe the page layout and any visible UI issues."

# Explicit provider
python vision.py --provider qwen "mockup.png" "List all components, colors, and spacing patterns."

# Custom model
QWEN_MODEL=qvq-max python vision.py --provider qwen "diagram.png" "Explain the architecture."

# GPT-4o for visual regression
python vision.py -p openai "after.png" "Compare with app design spec, flag differences."