benchmark evidence
UGI — ugi-overall
UGI Leaderboard — ugi-overall score from DontPlanToEnd/UGI-Leaderboard
activeupstream source ->
winner on UGI — ugi-overall
Mistral Large59.2
direct benchmark result, not a broad vertical composite | source row dated 2025-09-30
scored on 2026-04-25 · stale source data (35d)
latest mapped results | top 20
| # | Model | Score | Evidence | Tested |
|---|---|---|---|---|
| 1 | Mistral Large | 59.2 | model-only independent_benchmark | 2025-09-30 |
| 2 | MoonshotAI: Kimi K2 Thinking | 52.7 | model-only independent_benchmark | 2025-11-07 |
| 3 | OpenAI: o3 | 52.1 | model-only independent_benchmark | 2025-09-24 |
| 4 | OpenAI: GPT-5.5 | 51.2 | model-only independent_benchmark | 2026-04-25 |
| 5 | DeepSeek: R1 | 51.0 | model-only independent_benchmark | 2025-09-21 |
| 6 | Google: Gemini 2.5 Pro | 47.8 | model-only independent_benchmark | 2025-09-09 |
| 7 | OpenAI: GPT-4.1 | 47.5 | model-only independent_benchmark | 2025-09-24 |
| 8 | OpenAI: GPT-5.4 | 47.0 | model-only independent_benchmark | 2026-03-05 |
| 9 | OpenAI: GPT-5.1 | 44.3 | model-only independent_benchmark | 2025-11-16 |
| 10 | OpenAI: o1 | 42.3 | model-only independent_benchmark | 2025-09-24 |
| 11 | OpenAI: GPT-5 | 42.1 | model-only independent_benchmark | 2025-09-24 |
| 12 | OpenAI: o4 Mini | 38.2 | model-only independent_benchmark | 2025-10-04 |
| 13 | Meta: Llama 3.3 70B Instruct | 37.9 | model-only independent_benchmark | 2025-09-10 |
| 14 | Qwen: Qwen3 235B A22B | 35.2 | model-only independent_benchmark | 2025-09-21 |
| 15 | Mistral: Mistral Nemo | 33.4 | model-only independent_benchmark | 2025-09-06 |
| 16 | Qwen: Qwen3 32B | 31.0 | model-only independent_benchmark | 2025-09-13 |
| 17 | OpenAI: GPT-5.2 | 31.0 | model-only independent_benchmark | 2026-03-04 |
| 18 | Qwen: Qwen3 30B A3B | 28.8 | model-only independent_benchmark | 2025-11-08 |
| 19 | Anthropic: Claude Sonnet 4 | 23.2 | model-only independent_benchmark | 2025-09-12 |
| 20 | Google: Gemma 3 27B | 20.4 | model-only independent_benchmark | 2025-09-06 |
what this result means
UGI Leaderboard — ugi-overall score from DontPlanToEnd/UGI-Leaderboard
This benchmark contributes direct public evidence. Read its scope before generalizing the result.
A win here is a win on UGI — ugi-overall. Broad task pages require independent corroboration before naming a general winner.
source record
category: uncensored
metric: accuracy
matched models: 20
latest source date: 2026-04-25
direction: higher is better