Qwen·released 2025-08-051 source
Not yet benchmarked
Claims drawn from cited facts, not live model generation.
Developed by Qwen and released in 2025, this model is compact at 4.02 billion parameters. That compact scale favors efficiency and lower cost over the greater headroom and expense of larger models.
4 cited facts
This model is not yet benchmarked on any tracked benchmark. It holds no current top score.
2 cited facts
With an average SOTA gap that is unknown, the model cannot be placed on a points-behind-the-leader scale. It holds the top score on none of its benchmarks, so there is no evidence of category leadership yet.
2 cited facts
2 facts cross-checked across data sources: 2 single-source.
Source facts, citations, and refresh stamp for this record.
Sources: huggingface_models
Qwen3-4B-Instruct-2507 is an AI model developed by Qwen, released 2025-08-05. Its benchmark record below tags every score as independently reproduced, self-reported, or unverified.