ByteDance: Doubao-vision-pro-32k API
A flagship multimodal model utilizing a sparse MoE architecture with 20B active parameters. It specializes in "Thinking Mode" for complex visual reasoning, high-precision OCR, and dynamic resolution handling, offering GPT-4o level performance for high-stakes enterprise visual tasks.
- Context window: 32,768 tokens
- Max output: 12,288 tokens
- Input: text, image, video
- Output: text
- Reasoning: Supported
- Tool calling: Supported
- File input: Supported
- Released: 2025-01-21
- Knowledge cutoff: 2025-08-31
Frequently Asked Questions
What is the context window of Doubao-vision-pro-32k?
Doubao-vision-pro-32k supports a context window of up to 32,768 tokens.
Does Doubao-vision-pro-32k support function calling?
Yes. Doubao-vision-pro-32k supports tool / function calling.
Does Doubao-vision-pro-32k support reasoning?
Yes. Doubao-vision-pro-32k is a reasoning-capable model.
What is the knowledge cutoff of Doubao-vision-pro-32k?
The knowledge cutoff of Doubao-vision-pro-32k is 2025-08-31.