L3-DARKEST-PLANET-16.5B: benchmark results
DavidAU's 16.5B Brainstorm 40x expansion of his Llama 3 Dark Planet 8B merge, tuned for vivid, uncensored creative writing and fiction. Provider: Other. Released 2024-10-11. Access: Open.
Unified ELO 1493 ± 1, rank #725 of 1392 rated models, from 10 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - IFEval | 62.31 | Score | 74.3 |
| Open LLM Leaderboard - BBH | 31.78 | Score | 58.4 |
| Open LLM Leaderboard - MMLU-Pro | 29.23 | Score | 56.1 |
| Open LLM Leaderboard - GPQA | 6.04 | Score | 50.7 |
| Open LLM Leaderboard - MATH Level 5 | 8.99 | Score | 44.1 |
| UGI - Willingness (W/10) | 4.5 | W/10 Score | 34.4 |
| Open LLM Leaderboard - MuSR | 7.25 | Score | 34 |
| UGI - Writing | 26.74 | Writing Score | 27.1 |
| UGI Leaderboard | 23.49 | UGI Score | 16.8 |
| UGI - Natural Intelligence | 13.51 | NatInt Score | 13 |
Interactive version: theaggregate.ai/model?slug=l3-darkest-planet-16-5b · How It Works · Data refreshed daily, snapshot 2026-09-05.