Model comparison
DeepSeek-V2.5 (Sep 2024) vs Yi-1.5-34B
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 30.2 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeekSummary
- They share 19 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 7 categories and Yi-1.5-34B in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V2.5 (Sep 2024) leads 34.2 to 14.8.
- The biggest single-benchmark swing is BigCodeBench Instruct: 48.6% for DeepSeek-V2.5 (Sep 2024) and 33.9% for Yi-1.5-34B.
Side by side
| DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B | |
|---|---|---|
| Provider | DeepSeek | 01.AI |
| Noometry Index | 37.0 | 30.2 |
| Released | 2024-09-06 | 2024-05-13 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 21 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
DeepSeek-V2.5 (Sep 2024): 31.3 (#274), Yi-1.5-34B: 31.9 (#267)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| BigCodeBench Instruct | 48.6% | 33.9% |
| LMArena Coding | 1310 | 1169 |
| BigCodeBench Complete | 53.2% | 43.8% |
| Aider Polyglot | 17.8% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 24.5 (#145), Yi-1.5-34B: 21.5 (#195)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1160 |
Math DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 33.3 (#177), Yi-1.5-34B: 25.3 (#248)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Math | 1289 | 1182 |
| MATH Level 5 | — | 25.5% |
Knowledge DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 34.2 (#190), Yi-1.5-34B: 14.8 (#290)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Expert | 1266 | 1144 |
| GPQA Diamond | — | 32% |
Multilingual DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 42.7 (#188), Yi-1.5-34B: 32.5 (#251)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Non-English | 1273 | 1121 |
| LMArena Chinese | 1317 | 1213 |
| LMArena French | 1288 | 1155 |
| LMArena German | 1258 | 1111 |
| LMArena Japanese | 1228 | 1021 |
| LMArena Korean | 1209 | 1005 |
| LMArena Russian | 1289 | 1091 |
| LMArena Spanish | 1248 | 1120 |
Instruction Following DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 67.7 (#189), Yi-1.5-34B: 59.4 (#252)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Instruction Following | 1280 | 1139 |
Long Context DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#169), Yi-1.5-34B: 34.6 (#243)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Longer Query | 1301 | 1144 |
Writing & Preference DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 50.0 (#182), Yi-1.5-34B: 37.6 (#252)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | Yi-1.5-34B |
|---|---|---|
| LMArena Text | 1294 | 1173 |
| LMArena Creative Writing | 1285 | 1135 |
| LMArena Multi-Turn | 1297 | 1153 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than Yi-1.5-34B?
DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 30.2 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or Yi-1.5-34B better for coding?
They score almost the same on coding (31.3 vs 31.9); test both on your own repository before choosing.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Yi-1.5-34B share?
19 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Yi-1.5-34B has 21.