Qwen3 VL 235B A22B Instruct
Qwen3-VL-235B-A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video. The Instruct model targets general vision-language use (VQA, document parsing, chart/table extraction, multilingual OCR).
Input
Output
Context
262K
Max Output
262K
Parameters
235B
Technical Specifications
Model TypeVision
Context Window262K
Max Output Tokens262K
Parameters235B
Training Cutoff—
Licenseapache-2.0
Capabilities
Input Modalities
textimage
Output Modalities
text
Features
function_callingjson_modereasoningstreamingtext_generationvision
Resources & Links
Estimated GPU Requirements for Qwen3 VL 235B A22B Instruct
Estimates based on INT8 quantization. Actual requirements vary by framework and configuration.
Related Tools
Data sourced from official provider APIs and documentation
Last updated: May 5, 2026
Start building with the right model.
From model selection to production, one platform, no fragmentation.