Qwen
Qwen: Qwen3 VL 32B Instruct
Descripción del proveedor (EN)
Qwen3-VL-32B-Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video. With 32 billion parameters, it combines deep visual perception with advanced text...
ActivoFuente oficial ↗
FICHA TÉCNICA
Datos normalizados
- Id
- qwen/qwen3-vl-32b-instruct
- Name
- Qwen: Qwen3 VL 32B Instruct
- Provider
- Qwen
- Family
- qwen3-vl
- Source
- openrouter
- Created Unix
- 1761231332
- Knowledge Cutoff
- —
- Context Length
- 131072
- Source Id
- qwen/qwen3-vl-32b-instruct
- Canonical Slug
- qwen/qwen3-vl-32b-instruct
- Provider Slug
- qwen
- Variant
- —
- Is Free
- False
- Max Completion Tokens
- 32768
- Modality
- text+image->text
- Tokenizer
- Qwen
- Instruct Type
- —
- Is Moderated
- False
- Supported Parameters
- ["frequency_penalty","logprobs","max_tokens","presence_penalty","response_format","seed","stop","structured_outputs","temperature","tool_choice","tools","top_k","top_logprobs","top_p"]
- Default Parameters
- {"temperature":0.7,"top_p":0.8,"top_k":20,"frequency_penalty":null,"presence_penalty":null,"repetition_penalty":1}
- Reasoning
- {}
- Expiration Date
- 2026-10-09
- Hugging Face Id
- Qwen/Qwen3-VL-32B-Instruct
- Links
- {"details":"/api/v1/models/qwen/qwen3-vl-32b-instruct/endpoints"}
- Supports Tools
- True
- Supports Structured Outputs
- True
- Supports Reasoning
- False
- Supports Vision
- True
- Supports Audio
- False
- Supports Video
- False
- Supports Files
- False