MLX Benchmark V2 - MLX LM LoRA — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 69.09 |
| 2 | Qwen 3.6 Max Preview | 66.67 |
| 3 | Gemini 3 Flash (Preview) | 65.45 |
| 4 | Gemma 4 26B A4B (IT) | 63.64 |
| 5 | DeepSeek V4 Pro | 60.87 |
| 6 | GPT-5.4 Nano | 56.36 |
| 7 | DeepSeek V4 Flash | 56.36 |
| 8 | Grok 4.1 Fast | 56.36 |
| 9 | Gemini 2.5 Flash Lite (Preview 09-2025) | 52.73 |
| 10 | Qwen 3.6 35B A3B | 47.27 |
| 11 | GPT-5 Nano | 40 |
| 12 | GLM-5.1 | 18.18 |
| 13 | Nemotron 3 Ultra | 14.55 |
| 14 | Kimi K2.5 | 3.64 |
| 15 | Kimi K2.6 | 1.85 |
Interactive version: theaggregate.ai/benchmark?slug=mlx-benchmark-v2-mlx-lm-lora · How the rankings work · Data refreshed daily, snapshot 2026-07-22.