Qwen3.8 Flash

Alibaba · Multimodal · Released Aug 2026

Qwen3.8 Flash is Alibaba's lightweight multimodal reasoning model supporting text, image, and video inputs with a 1 million token context window.

Strengths: Handles coding assistance, visual understanding, document analysis, and long-video processing while maintaining fast inference throughput.
Best for: Agentic workflows, desktop interaction, chart analysis, and codebase reasoning where multimodal input and reasoning speed matter.
Limitations: A faster tier compared to Qwen3.8 Max; trade-offs in reasoning depth and parameter count are typical for models optimized for latency.

Input / 1M

$0.15

Output / 1M

$0.47

Cached input / 1M

$0.016

Context window

Also billed

Price history

Effective	Input	Output	Cached in	Note	Source
27 Aug 2026	$0.15	$0.47	$0.016	Imported from OpenRouter	openrouter.ai

Qwen3.8 27B

in $0.42 · out $3.00

Qwen3.8 2.4T A95B

in $2.00 · out $6.00

Qwen3.8 2.4T A95B (batch)

in $2.00 · out $6.00

Qwen3.8 Max

in $2.00 · out $6.00

Data updated Aug 27, 2026 Report a problem