Qwen3-Coder-480B-A35B

Training compute

1.6×10²⁴ FLOP

Parameters

480B

Published

Jul 22, 2025

Qwen3-Coder-480B-A35B is an AI model developed by Alibaba (China), first published in July 2025. It works in the language domain, on tasks such as language modeling/generation, question answering, code generation and system control.

Training it took an estimated 1.6×10²⁴ FLOP of compute (estimation method: operation counting). The model has 480,000,000,000 parameters. It was trained on roughly 7.5T datapoints.

Access: Open weights (unrestricted). Its weights are openly available. Epoch AI rates the confidence of this record as confident.

Full record

Organization: Alibaba
Country of organization: China
Domain: Language
Task: Language modeling/generation, Question answering, Code generation, System control
Training compute: 1.6×10²⁴ FLOP
Compute estimation method: Operation counting
Parameters: 480,000,000,000
Dataset size: 7.5T
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Confident