Model comparison
DeepSeek-V2.5 (Sep 2024) vs ERNIE 5.0 0110
ERNIE 5.0 0110 is the stronger model overall, scoring 41.8 to 37.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
BaiduSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 1 category and ERNIE 5.0 0110 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where ERNIE 5.0 0110 leads 63.1 to 49.8.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 | |
|---|---|---|
| Provider | DeepSeek | Baidu |
| Noometry Index | 37.6 | 41.8 |
| Released | 2024-09-06 | — |
| Weights | Open | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 22 | 20 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), ERNIE 5.0 0110: 43.0 (#94)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Coding | 1309 | 1455 |
| Aider Polyglot | 17.8% | — |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), ERNIE 5.0 0110: 17.0 (#297)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1445 |
| NYT Connections (extended) | — | 10.3% |
| Thematic Generalization | — | 41.7% |
Math ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), ERNIE 5.0 0110: 39.3 (#110)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Math | 1288 | 1437 |
Knowledge ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), ERNIE 5.0 0110: 39.8 (#128)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Expert | 1266 | 1428 |
Multimodal Not comparable
DeepSeek-V2.5 (Sep 2024): —, ERNIE 5.0 0110: 39.9 (#53)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Vision | — | 1249 |
Multilingual ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), ERNIE 5.0 0110: 54.1 (#49)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Non-English | 1273 | 1436 |
| LMArena Chinese | 1318 | 1512 |
| LMArena French | 1289 | 1467 |
| LMArena German | 1258 | 1460 |
| LMArena Japanese | 1228 | 1382 |
| LMArena Korean | 1209 | 1406 |
| LMArena Russian | 1289 | 1446 |
| LMArena Spanish | 1248 | 1473 |
Instruction Following ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), ERNIE 5.0 0110: 74.5 (#92)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Instruction Following | 1280 | 1413 |
Long Context ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), ERNIE 5.0 0110: 43.4 (#95)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Longer Query | 1301 | 1422 |
Writing & Preference ERNIE 5.0 0110 leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), ERNIE 5.0 0110: 63.1 (#66)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | ERNIE 5.0 0110 |
|---|---|---|
| LMArena Text | 1294 | 1445 |
| LMArena Creative Writing | 1285 | 1426 |
| LMArena Multi-Turn | 1297 | 1434 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than ERNIE 5.0 0110?
ERNIE 5.0 0110 is the stronger model overall, scoring 41.8 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or ERNIE 5.0 0110 better for coding?
ERNIE 5.0 0110 scores higher on coding benchmarks: 43.0 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and ERNIE 5.0 0110 share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and ERNIE 5.0 0110 has 20.