← Back to leaderboard
Head-to-head
Code Llama 34B Instruct vs openai/gpt-oss-20b
openai/gpt-oss-20b wins+27.7 pts (+43%)on the balanced composite.
Category-by-category
Higher bar wins each row.
Code Llama 34B Instruct
Category
openai/gpt-oss-20b
-35.364.7
Code
100.0+35.3
-10.264.8
Reason
75.0+10.2
-19.468.1
Tool Use
87.5+19.4
-33.866.2
RAG
100.0+33.8
-39.060.3
Speed
99.3+39.0
Code Llama 34B Instruct
- Provider
- meta
- Released
- 2024-01-29
- Context
- 16.384K
- Lab
- Community
openai/gpt-oss-20b
- Provider
- local
- Released
- Context
- 0K
- Lab
- Community