Command Palette
Search for a command to run

Qwen3.5 122B A10B

by Qwen

Specifications

Input
Output
Context window
262K tokens
Released
Feb 2026

Performance

Speed
38 t/s
TTFT
106 ms
Latency
Intelligence

Pricing

Input
€0.24
per 1M tokens
Output
€1.91
per 1M tokens

About this model

Qwen3.5-122B-A10B is Alibaba Cloud's native multimodal agent model with 122B total parameters (10B activated). Features 240K context, vision capabilities, hybrid reasoning with extended thinking, function calling, and support for 201 languages. Apache 2.0 licensed.

Technical specifications

Capabilities
Input modalities
Output modalities
Reasoning
Hybrid Default on

Knowledge horizon

Knowledge cutoff Sep 2025
Released Feb 2026
Today
Training to release 5 mo Since release 3 mo

See also

Add Model to Comparison
Search for a model to add
Command Palette
Search for a command to run