Qwen: Qwen3.5 397B A17B
Qwen
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers...
Pricing
Source-reported prices are shown in their native units. No missing fee is estimated.
Input
$0.500
USD / 1M input tokens
Output
$3.60
USD / 1M output tokens
Capabilities
A check means the source explicitly lists support; a dash means the source did not publish a value.
Text input Supported
Text output Supported
Vision Supported
Tools Supported
Reasoning Supported
Structured outputs Supported
Audio Not listed
Image generation Not listed
Embeddings Not listed
Alternatives
Suggestions use deterministic rules: overlapping modalities, a context window within 2×, and shared source-reported signals. No AI-generated recommendations.
Similar models
AionLabs: Aion 3.5 Mini
aion-labs/aion-3.5-mini
Input
$0.700
Intelligence
N/A
Similar category, context, price or shared benchmark signals.
AionLabs: Aion-2.0
aion-labs/aion-2.0
Input
$0.800
Intelligence
N/A
Similar category, context, price or shared benchmark signals.
AionLabs: Aion-3.0-Mini
aion-labs/aion-3.0-mini
Input
$0.700
Intelligence
N/A
Similar category, context, price or shared benchmark signals.
Amazon: Nova Pro 1.0
amazon/nova-pro-v1
Input
$0.800
Intelligence
N/A
Similar category, context, price or shared benchmark signals.