Hub/Models/qwen/qwen3-235b-a22b-2507
qwentext → text

Qwen: Qwen3 235B A22B Instruct 2507

Canonical ID: qwen/qwen3-235b-a22b-2507

Compare ModelOpen on OpenRouter →

Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following,...

Artificial Analysis Benchmarks

Verified
No standardized Artificial Analysis benchmark scores recorded for this variant yet.

Token Pricing & Economics

Prompt Input (1M Tokens)
$0.14
Input
Completion Output (1M Tokens)
$0.80
Output
Prompt KV-Cache Read (1M Tokens)
$0.05
Cached
Context Window Capacity
256k tokens
Context

Supported Parameters & Features

frequency_penaltylogit_biaslogprobsmax_tokensmin_ppresence_penaltyrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p