← Back to leaderboard
Head-to-head
mlx-community/Qwen3.6-35B-A3B-4bit vs openai/gpt-oss-20b
openai/gpt-oss-20b wins+8.3 pts (+10%)on the balanced composite.
Category-by-category
Higher bar wins each row.
mlx-community/Qwen3.6-35B-A3B-4bit
Category
openai/gpt-oss-20b
-25.075.0
Code
100.0+25.0
+5.080.0
Reason
75.0-5.0
87.5
Tool Use
87.5
100.0
RAG
100.0
-16.782.5
Speed
99.3+16.7
mlx-community/Qwen3.6-35B-A3B-4bit
- Provider
- local
- Released
- Context
- 0K
- Lab
- Community
openai/gpt-oss-20b
- Provider
- local
- Released
- Context
- 0K
- Lab
- Community