qwenBerbayar
Qwen: Qwen3 235B A22B Instruct 2507
qwen/qwen3-235b-a22b-2507
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,...
Harga input / 1M token
$0.0875
Harga output / 1M token
$0.3500
Context length
262.144
Modalitas
- Input
- text
- Output
- text
Parameter yang didukung
frequency_penaltylogit_biaslogprobsmax_tokensmin_ppresence_penaltyrepetition_penaltyresponse_formatseedstopstructured_outputstemperaturetool_choicetoolstop_ktop_logprobstop_p
Rilis
21 Juli 2025