Qwen: Qwen3 VL 8B Thinking
Released Oct 14, 2025Try it on:
Qwen3-VL-8B-Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences. It integrates enhanced multimodal alignment and...
Best overall rank
-
Across all subsets · overall
Total arena votes
0
Across 0 (subset, category) entries
Providers on OR
0
None yet
First seen
-
Awaiting first snapshot
Trend
No snapshot history yet - trends appear after the first daily refresh.
Pricing
Per token
- Input (per 1M)
- $0.117
- Output (per 1M)
- $1.36
Capabilities
Modality
text+image->textContext
256K tokens
Tokenizer
Qwen3
Max output
33K tokens
Moderation
Unmoderated
Hugging Face
Supported parameters
include_reasoning
max_tokens
presence_penalty
reasoning
response_format
seed
structured_outputs
temperature
tool_choice
tools
top_p
Providers
No provider endpoint data yet.
Per-category ranks
No leaderboard entries for this model.