EffiBench - NET: leaderboard
EffiBench normalized execution time (NET) — runtime efficiency of model-generated programs relative to reference solutions; lower is better.
Metric: Normalized Execution Time. Source: huggingface.co. 42 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | starcoder2-15B | 2.59 | #1495 |
| 2 | Yi Chat 34B | 2.77 | #976 |
| 3 | Yi 34B (Base) | 2.81 | #972 |
| 4 | deepseek-coder-1.3B-instruct | 2.82 | #1585 |
| 5 | starcoder | 2.84 | #1577 |
| 6 | WizardCoder-15B-V1.0 | 2.84 | #1588 |
| 7 | deepseek-coder-6.7B-instruct | 2.89 | #1335 |
| 8 | starcoder2-3B | 2.9 | #1596 |
| 9 | Yi-34B-200K | 2.91 | #1176 |
| 10 | CodeLlama-13B-Instruct-hf | 2.93 | #1532 |
| 11 | CodeLlama-7B-hf | 2.95 | #1457 |
| 12 | CodeLlama-34B-hf | 2.98 | #1515 |
| 13 | deepseek-coder-6.7B-base | 2.98 | #1484 |
| 14 | Magicoder-S-DS-6.7B | 2.99 | #1256 |
| 15 | starcoder2-7B | 3.02 | #1572 |
Interactive version: theaggregate.ai/benchmark?slug=effibench-net · How It Works · Data refreshed daily, snapshot 2026-10-20.