whatmodel.sh / Yi-Large
Yi-Large
↗ share
+ contribute
Yi-Large is a closed-weights language model from 01.AI, released May 2024. Licence: proprietary.
Catalogue identity notes
Comparison-only. 01.AI's proprietary flagship API model 'yi-large' (launched with Yi-1.5 in May 2024; released month from the evaluation record, not a vendor page: 01.AI's international API platform is gone). LMArena's leaderboard table (arena-leaderboard Space @6f2eeeb) lists key 'yi-large' as 'Yi-Large' (01 AI, Proprietary) linking https://platform.01.ai/docs#models-and-pricing. The id is undated and no 01.AI page documents a version change; results come from ZeroEval, WildBench v2 and Arena-Hard v0.1 runs labelled 'yi-large' (2024). Distinct from the earlier 'yi-large-preview'.
$ how it compares
The LLMs ranked either side of Yi-Large, out of 828 with a score, then more from 01.AI.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 394 |  Ministral 3 3B Reasoning 2512 | 51±1 | local | ||
| 395 |  Qwen3 4B | 51±1 | local | ||
| 396 | Yi-Large | 51±1 | comparison only | ||
| 397 | Llama-3-Instruct-8B-SimPO-ExPO | 51±1 | local | ||
| 398 |  LFM2 24B-A2B | 51±1 | local | ||
| more from 01.AI | |||||
| 308 | Yi-Lightning | 55±1 | api price unknown | ||
| 373 | Yi-Large Preview | 52±1 | comparison only | ||
| 414 | Yi 34B 200K (base) | 50±5 | local | ||
$ what it's good at
coding▁▁▁▃▅▇█▇▅▄▃▁ #306 of 524 * · 20% of its score · 1 test
Beats Qwen3-VL 32B Instruct, Solar Open 100B (102B-A12B), GLM-4.7-Flash here, though they rank higher overall. Trails DeepSeek LLM 67B Chat, QwQ-32B-Preview, Phi-3 Medium 128K Instruct here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| CRUXEval (ZeroEval, 800 problems) | 60.3% ✎ | ▁▂▆▃▆█▆▃▂·▁▁ beats 20/47 | 20% |
chat & writing▁▁▁▃▅▄▆█▇▇▇▃ #264 of 463 * · 20% of its score · 1 test
Beats Granite 4.2 8B, Mistral Small 4 (119B-A6.5B), Olmo 3.1 32B Think here, though they rank higher overall. Trails Flan-T5 XXL, LFM2.5 1.2B Instruct, Gemma 3 4B here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena-Hard (v0.1, GPT-4-1106 judge) | 63.7% ✎ | ▂▁▁▅█▅▄▃▃▅▃▁ beats 43/60 | 20% |
math▁▂▃▅▇▅▅█▄▁▁▁ #395 of 462 * · 7.8% of its score · 1 test
Trails Qwen3 0.6B, Gemma 7B (base), LFM2 700M here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| GSM8K (ZeroEval, zero-shot) | 80.1% ✎ | ▁▄·▁▄█▅▅▅███ beats 9/43 | 7.8% |
other tests3 not in a task · 52% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| MMLU-Redux (ZeroEval, 2,778 questions) | 81.2% ✎ | ▁·▁▁▄▄▄▅█▄▂▁ beats 31/53 | 31% |
| ZebraLogic (ZeroEval, 1,000 puzzles) | 18.8% ✎ | ▁▁▃▆██▁▁·▁▁▁ beats 26/57 | 12% |
| WildBench v2 WB-Score (task macro) | 48.9273 WB-Score ✎ | ▁·▁▁·▃█▄▇▆▅▆ beats 40/52 | 8.3% |
$ where to run it
No verified API or download recorded yet · add one
$ how the score was worked out · WildEval ZeroEval published r… 71%, Arena-Hard-Auto v0.1 leaderbo… 20%, WildBench v2 leaderboard data… 8%
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 WildEval ZeroEval published results | 4 | 71% |
 Arena-Hard-Auto v0.1 leaderboard (repository README, pre-v2) | 1 | 20% |
| WildBench v2 leaderboard data (Ai2) | 1 | 8% |
| prior · pull toward the middle | 0% | |
| weighted average | = 50.9 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json