Magician-MoE-4x7B: benchmark results
A four-expert mixture assembled by FelixChao from DeepSeek Coder, Magicoder, WizardMath and WizardCoder checkpoints, with cheap_embed gating and model-specific routing prompts. Provider: FelixChao. Access: Open.
Aggregate Rating 1099 ± 39, rank #10729 of 10916 rated models, from 5 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard v1 - ARC Challenge | 28.24 | Normalized accuracy (%) (25-shot) | 7.5 |
| Open LLM Leaderboard v1 - HellaSwag | 30.06 | Normalized accuracy (%) (10-shot) | 5.4 |
| Open LLM Leaderboard v1 - MMLU | 24.67 | Accuracy (%) (5-shot) | 4.7 |
| Open LLM Leaderboard v1 - GSM8K | 0 | Accuracy (%) (5-shot) | 4.5 |
| Open LLM Leaderboard v1 - WinoGrande | 49.88 | Accuracy (%) (5-shot) | 2.6 |
Interactive version: theaggregate.ai/model?slug=magician-moe-4x7b · How It Works · Data refreshed daily, snapshot 2026-11-08.