DarkForest-20B-v2.0: benchmark results
A two-stage storytelling merge released by TeeZee: first Orca-2-13b and LLaMA2-13B-Erebus-v3 produce an expanded model, then psyonic-cetacean-20B and BigMaid-20B-v1.0 contribute to the final release. Provider: TeeZee. Access: Open.
Aggregate Rating 1394 ± 35, rank #6366 of 10916 rated models, from 6 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - HellaSwag | 86.32 | Normalized accuracy (%) (10-shot) | 80.2 |
| Open LLM Leaderboard v1 - TruthfulQA MC2 | 56.14 | MC2 (%) (0-shot) | 67.1 |
| Open LLM Leaderboard v1 - ARC Challenge | 63.74 | Normalized accuracy (%) (25-shot) | 64.1 |
| Open LLM Leaderboard v1 - WinoGrande | 77.9 | Accuracy (%) (5-shot) | 59.3 |
| Open LLM Leaderboard v1 - MMLU | 59.79 | Accuracy (%) (5-shot) | 45.9 |
| Open LLM Leaderboard v1 - GSM8K | 23.28 | Accuracy (%) (5-shot) | 42.9 |
Interactive version: theaggregate.ai/model?slug=darkforest-20b-v2-0 · How It Works · Data refreshed daily, snapshot 2026-11-08.