Mmodel.kanadakurniwan
qwenBerbayar

Qwen: Qwen3.5-122B-A10B

qwen/qwen3.5-122b-a10b

The Qwen3.5 122B-A10B 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. In terms of...

Harga input / 1M token
$0.2900
Harga output / 1M token
$2.40
Context length
262.144

Modalitas

Input
text, image, video
Output
text

Parameter yang didukung

frequency_penaltyinclude_reasoninglogit_biaslogprobsmax_tokensmin_ppresence_penaltyreasoningrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p

Rilis

25 Februari 2026

Buka di OpenRouter ↗