LOADING COMPARISON
GROUNDED IN REAL HUMAN CONVERSATIONSYOUR BENCH
PICK MODELS TO COMPARE
Now pick a second model to put against DeepSeek R1 Qwen3 8B. Fair fights only, so these are all chat and coding models.
SELECTED: DEEPSEEK R1 QWEN3 8B
Grok 4.6MULTIMODAL#1 · 41,353 VOICES
7.7VIBE
Qwen3.8 27BMULTIMODAL#2 · 13,889 VOICES
7.6VIBE- DiscoveryTEXT
#3 · 9,505 VOICES
7.6VIBE 
GLM 5.3MULTIMODAL#4 · 9,726 VOICES
7.5VIBE- Step 2TEXT
#7 · 3,680 VOICES
7.5VIBE 
DeepSeek V4 FlashTEXT#10 · 6,204 VOICES
7.4VIBE- Step 1TEXT
#11 · 3,692 VOICES
7.4VIBE - Dots3 NoteMULTIMODAL
#12 · 2,634 VOICES
7.4VIBE - Apertus v1.5 70BTEXT
#13 · 2,526 VOICES
7.4VIBE - UncensoredTEXT
#17 · 2,176 VOICES
7.3VIBE - CommunityTEXT
#18 · 1,854 VOICES
7.3VIBE 
Gemini 3.7 FlashMULTIMODAL#20 · 6,764 VOICES
7.2VIBE