Qwen: Qwen3 VL 8B Thinking

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...

• Context window: 256,000 tokens
• Input: image, text
• Output: text
• Pricing: $0.117/M input tokens, $1.365/M output tokens

View on OpenRouter. Model data sourced from OpenRouter.