Head-to-head
Compare model configs
Pick two model × quant rows and inspect quality, effective VRAM, speed, and per-axis winners side by side. Ranks only compare within the same lane.
Left
Gemma 4 31B IT
Q4_K_M
- Local Intelligence IndexLB-2026-07.2 | 25/22.5/22.5/22.5/7.5
- 51.7full index
- Agentic 10.4 / Knowledge 87.4 / Instruction 79.3 / Coding 35.5 / Math 48.2
- Effective VRAM
- 20.9 GB
- Fits
- 24 GB
- tok/s
- 48.4
Right
Qwen3.6 27B
Q4_K_M
- Local Intelligence IndexLB-2026-07.2 | 25/22.5/22.5/22.5/7.5
- 43.2full index
- Agentic 8.3 / Knowledge 83.4 / Instruction 65.0 / Coding 25.5 / Math 26.6
- Effective VRAM
- 19.5 GB
- Fits
- 24 GB
- tok/s
- 69.4
Local Intelligence Index delta
LB-2026-07.2 | 25/22.5/22.5/22.5/7.5. Profile: Agentic / Knowledge / Instruction / Coding / Math deltas appear below.
+8.5
VRAM delta
1.4 GB
tok/s delta
-21.0
Swipe horizontally for per-axis deltas →
| Axis | Left | Right | Delta | Winner |
|---|---|---|---|---|
| Agentic | 10.4 | 8.3 | +2.1 | Gemma 4 31B IT Q4_K_M wins |
| Knowledge | 87.4 | 83.4 | +4.0 | Gemma 4 31B IT Q4_K_M wins |
| Instruction | 79.3 | 65.0 | +14.3 | Gemma 4 31B IT Q4_K_M wins |
| Coding | 35.5 | 25.5 | +10.0 | Gemma 4 31B IT Q4_K_M wins |
| Math | 48.2 | 26.6 | +21.6 | Gemma 4 31B IT Q4_K_M wins |