#19Alibaba QwenOpen weights
Qwen3.8 27B
Alibaba's dense 27B open-weights reasoning model. Scores 34 on the Artificial Analysis Intelligence Index - the highest of any model under 40B parameters. Apache 2.0 licensed.
- Overall
- 52.3
- Input $/M
- $0.50
- Output $/M
- $3.00
- Context
- 262K
- Released
- Aug 14, 2026
- Updated
- Oct 3, 2026
Category scores
- Reasoning & Mathrank 1858.6
- Speedrank 2018.4
- Cost-efficiencyrank 961.2
- Open-sourcerank 558.6
- Best small modelrank 158.6
Overall rank history
Weekly snapshots. Lower is better.
Strengths
- Best small model by a clear margin (score 58.6)
- Open weights (Apache 2.0)
- Good cost-efficiency (61.2)
Weaknesses
- Slow generation (speed score 18.4)
- Higher output price ($3.00 per 1M) than small-model rivals