LOADING COMPARISON
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
PICK MODELS TO COMPARE
Now pick a second model to put against Qwen2.5 Coder 32B Instruct. Fair fights only, so these are all chat and coding models.
SELECTED: QWEN2.5 CODER 32B INSTRUCT
Grok 4.6MULTIMODAL#1 · 42,704 VOICES
7.7VIBE
Qwen3.8 27BMULTIMODAL#2 · 14,071 VOICES
7.6VIBE- DiscoveryTEXT
#3 · 10,083 VOICES
7.6VIBE - Qwen3.8 FlashMULTIMODAL
#4 · 6,013 VOICES
7.6VIBE 
GLM 5.3MULTIMODAL#5 · 9,898 VOICES
7.5VIBE- Step 2TEXT
#8 · 3,787 VOICES
7.5VIBE - GLM 5.3 FlashMULTIMODAL
#10 · 6,992 VOICES
7.4VIBE 
DeepSeek V4 FlashTEXT#11 · 6,461 VOICES
7.4VIBE- Step 1TEXT
#13 · 3,811 VOICES
7.4VIBE - Dots3 NoteMULTIMODAL
#14 · 2,634 VOICES
7.4VIBE - Apertus v1.5 70BTEXT
#15 · 2,526 VOICES
7.4VIBE - CommunityTEXT
#16 · 2,335 VOICES
7.4VIBE