01.AI
Yi-34B benchmark results
8 results on 6 benchmarks from 4 sources.
Open weights Released: Nov 2, 2023 Updated Oct 23, 2026
Also listed as: Yi 34B Chat · Yi-34B-Chat · Yi (34B) · Yi 34B · Yi-Chat 34B · 01-ai/Yi-34B-Chat · 01-ai/Yi-34B · 01-ai/Yi-34B-200K
All benchmark results
| Benchmarks | Score | Rank | Setting | Source type | Date |
|---|---|---|---|---|---|
| LMArena Text | 1184 | 357 | Overall, style control on; 15483 votes; 95% CI 1177–1191“Yi 34B Chat” | Official leaderboard | Oct 22, 2026 |
| Open LLM Leaderboard v2 | 24.2% | 1759 | Average of 6 normalised benchmarks; bfloat16; 34.389B params“01-ai/Yi-34B-Chat” | Official leaderboard | Mar 13, 2025 |
| GPQA Diamond | 14.7% | 280 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
| MMLU | 76.3% | 17 | Exact match; 5-shot | Paper / tech report via Epoch AI | Nov 2, 2023 |
| MMLU | 67.6% | 33 | Exact match; 0-shot | Paper / tech report via Epoch AI | Nov 22, 2023 |
| MMLU | 76.2% | 36 | Exact match; 5-shot | Official leaderboard via Epoch AI | Nov 2, 2023 |
| MMLU-Pro | 43.0% | 184 | Overall accuracy; data source: TIGER-Lab; 34B | Official leaderboard | Mar 11, 2026 |
| MATH Level 5 | 5.2% | 102 | Epoch AI run, accuracy | Independent run | Jan 27, 2025 |
Where Yi-34B ranks
Primary setting, one best-sourced score per model (official / independent first); neighbours in brackets.
| Benchmark | Score | Rank | Nearby models | #1 |
|---|---|---|---|---|
| MMLU | 76.2% | #44 / 158 | Qwen1.5 Chat (110B) 76.8% · GPT-4o mini 76.7% · Claude 3 Sonnet 75.9% · phi-3-small 7.4B 75.7% | GPT-4o (Nov 2024)88.1% |
| Open LLM Leaderboard v2 | 24.2% | #83 / 294 | Falcon3-7B-Base 24.8% · Mistral NeMo 24.7% · Llama-3.2-3B-Instruct 24.2% · Ministral 8B 24.2% | Qwen2.5-72B48.0% |
| MMLU-Pro | 43.0% | #180 / 243 | MAmmoTH2-8B-Plus 43.4% · Mixtral 8x7B 43.3% · Claude 3 Haiku 42.3% · Mathstral-7B-v0.1 42.0% | Gemini 3.1 Pro91.2% |
| LMArena Text | 1184 | #332 / 392 | DeepSeek LLM 67B 1185 · WizardLM 70B 1185 · Granite 3.0 8B Instruct 1183 · OpenChat 3.5 1183 | Gemini 4 Argon1517 |
| MATH Level 5 | 5.2% | #92 / 94 | DeepSeek LLM 67B 6.4% · Llama 3-8B 6.1% · Mistral 7B v0.3 3.6% · Llama 2-70B 3.3% | GPT-598.1% |
| GPQA Diamond | 14.7% | #215 / 217 | Gemma 3 1B 19.9% · Mistral 7B v0.3 15.2% · granite-4.0-350m 11.2% · deepseek-r1-0528-qwen3-8b 9.3% | Gemini 4 Argon96.4% |
Other 01.AI models
FAQ
Where does Yi-34B rank best?
MMLU: #44 of 158 (76.2%); #1 there is GPT-4o (Nov 2024) (88.1%).
Where do these Yi-34B scores come from?
From Epoch AI Benchmarking Hub, LMArena leaderboard dataset, MMLU-Pro leaderboard (TIGER-Lab), and Hugging Face Open LLM Leaderboard v2 (archived). Each row links to the original leaderboard, dataset or report, and vendor-reported numbers are labelled.