Public results
Leaderboard
Every score is uploaded from a Model Arena desktop install. Filter by category, search models, set VRAM and context ranges, and add models to compare them side by side.
Models
6
Tests
15
8 categories
Submissions
365
Updated
46d ago
Filter
No categories selected = every scored testVRAM–GBCtx–k
Runner / Provider
Average score per model
6 models · mean of per-test averagesPlacements per test
Each dot = one model's average on one testCoverage
Color = score (red 0 → lime 10) · faded = outside current filter · gray = not uploaded0
10
Ranking
Click a column to sort · click a runner pill to filter · + adds to compare (max 5)| 1 | DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF 27B· 18GB | Q4_K_M | 2.3 GB | 50k | 6.9 | 6.5 | 9.5 | 10.0 | 7.3 | 9.0 | 8.1/10 n=15 | ||
| 2 | ISTA-DASLab/Qwen3.8-27B-GSQ-RCO-GGUF 27B· 12GB | IQ3_S | 19.8 GB | 75k | 8.0 | 7.3 | 10.0 | 9.0 | 6.0 | 7.3 | 8.0/10 n=11 | ||
| 3 | unsloth/gemma-4-E2B-it-qat-GGUF 2B· 2.6GB | UD-Q4_K_XL | 4 GB | 131k | 6.0 | 3.0 | 7.5 | 9.0 | 8.0 | 8.8 | 7.3/10 n=10 | ||
| 4 | OBLITERATUS/Qwen3.8-27B-OBLITERATED 27B· 17GB | Q4_K_M | 22.1 GB | 100k | — | 0.0 | 4.3 | 10.0 | 9.3 | 9.0 | 6.6/10 n=8 | ||
| 5 | unsloth/LFM2.5-8B-A1B-GGUF 8B· 9.3GB | UD-Q8_K_XL | 13.2 GB | 125k | 7.0 | 3.0 | 7.0 | 0.0 | 4.7 | 7.0 | 6.4/10 n=9 | ||
| 6 | unsloth/gpt-oss-20b-GGUF 20b· 14GB | F16 | 16.6 GB | 131k | 6.2 | 3.8 | 4.8 | 8.5 | 5.9 | 8.0 | 6.2/10 n=14 |