PipelineScore
← Back to leaderboard
Head-to-head

/home/thomaskyn/models/Qwen3.5-9b/Qwen3.5-9B-Q4_K_M.gguf vs DeepSeek V3 671B-A37B

DeepSeek V3 671B-A37B wins+31.5 pts (+62%)on the balanced composite.

Category-by-category

Higher bar wins each row.

/home/thomaskyn/models/Qwen3.5-9b/Qwen3.5-9B-Q4_K_M.gguf
Category
DeepSeek V3 671B-A37B
-56.325.0
Code
81.3+56.3
-22.360.0
Reason
82.3+22.3
+3.387.5
Tool Use
84.3-3.3
-10.175.0
RAG
85.1+10.1
-67.712.0
Speed
79.7+67.7
/home/thomaskyn/models/Qwen3.5-9b/Qwen3.5-9B-Q4_K_M.gguf
Provider
local
Released
Context
0K
Lab
Community
DeepSeek V3 671B-A37B
Provider
deepseek
Released
2024-12-26
Context
128K
Lab
Community