Yi 34B (Base) — benchmark results
01.AI's original 34B bilingual English/Chinese base model (November 2023) that topped the Open LLM Leaderboard for pretrained models at release. Provider: 01.AI. Released 2023-11-02. Access: Open.
Unified ELO 1419 ± 19, rank #1135 of 1776 rated models, from 51 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Chinese LLM - WinoGrande | 70.8 | Accuracy (%) | 95.1 |
| Open LLM Leaderboard - GPQA | 15.55 | Score | 91.9 |
| Open Chinese LLM - HellaSwag | 68.92 | Accuracy (%) | 89.6 |
| HELM NarrativeQA | 78.22 | F1 (%) | 88.9 |
| URIAL-Bench - Reasoning | 6 | Judge Score (0-10) | 88.9 |
| EffiBench - NET | 2.81 | Normalized Execution Time | 87.8 |
| EffiBench - NMU | 1.89 | Normalized Memory Usage | 87.8 |
| HELM NaturalQuestions (Open) | 77.51 | F1 (%) | 86.7 |
| Open Chinese LLM - C-Eval Semantic | 87.83 | Accuracy (%) | 86.1 |
| URIAL-Bench - Roleplay | 7.75 | Judge Score (0-10) | 83.3 |
| Open Chinese LLM - CMMLU | 69.79 | Accuracy (%) | 83.1 |
| HELM v2 Lite - HumanEval (Code) | 30 | Pass@1 (%) | 82.4 |
Interactive version: theaggregate.ai/model?slug=yi-34b-base · How the rankings work · Data refreshed daily, snapshot 2026-07-22.