prices synced 2026-07-28

Qwen3 VL 235B A22B Instruct

Alibaba · Multimodal · Released Sep 2025

Compare

Qwen3 VL 235B A22B Instruct is a 235-billion-parameter multimodal vision-language model from Alibaba that processes text, images, and video within a 131k-token context window.

Strengths
The model's large parameter count and multimodal capabilities enable it to handle complex visual reasoning and understand context across extended sequences.
Best for
Dense visual analysis tasks, multimodal document understanding, and long-context problems that benefit from a high-capacity model.
Limitations
The 131k-token context window is smaller than some other Alibaba models like Qwen3 VL 30B A3B Instruct (262k tokens), and this instruct variant lacks the extended reasoning capabilities of the concurrent Qwen3 VL 235B A22B Thinking release.

Input / 1M

$0.21

Output / 1M

$1.90

Cached input / 1M

$0.1

Context window

131K

Price history

Input (solid)Output (dashed)
Price per 1M tokens over timeOutput went from $0.88 to $1.90; Input went from $0.2 to $0.21 between Jun 2026 and Jul 2026.$0$0.5$1.00$1.50$2.00Output on 11 Jun 2026: $0.88Output on 15 Jul 2026: $1.90Out $1.90Input on 11 Jun 2026: $0.2Input on 15 Jul 2026: $0.21In $0.21Jun 2026Jul 2026

Price change

30d
in increased 5.0% out increased 115.9%
90d
in increased 5.0% out increased 115.9%
1y
in increased 5.0% out increased 115.9%
Since launch
in increased 5.0% out increased 115.9%

Snapshots

Effective Input Output Cached in Note Source
15 Jul 2026 $0.21 $1.90 $0.1 Imported from OpenRouter openrouter.ai
11 Jun 2026 $0.2 $0.88 $0.11 Imported from OpenRouter openrouter.ai

More from Alibaba

Report a problem