Model comparison
ERNIE 5.1 vs Qwen2.5 72B Instruct
ERNIE 5.1 is the stronger model overall, scoring 43.8 to 31.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
Baidu
Alibaba (Qwen)Summary
- They share 17 benchmarks with published results for both. ERNIE 5.1 scores higher in 7 categories and Qwen2.5 72B Instruct in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where ERNIE 5.1 leads 40.3 to 19.3.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| ERNIE 5.1 | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | Baidu | Alibaba (Qwen) |
| Noometry Index | 43.8 | 31.9 |
| Released | — | 2024-09 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 8K |
| Input $ / M tokens | — | $1.40 |
| Output $ / M tokens | — | $5.60 |
| Results tracked | 19 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding ERNIE 5.1 leads
ERNIE 5.1: 44.0 (#76), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Coding | 1488 | 1292 |
| WeirdML | — | 16% |
| BigCodeBench Instruct | — | 45.8% |
| BigCodeBench Complete | — | 55.9% |
Agentic & Tool Use Not comparable
ERNIE 5.1: —, Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| LMArena Search | 1227 | — |
| METR Time Horizons | — | 35.8% |
Reasoning Too close to call
ERNIE 5.1: 21.9 (#211), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1481 | 1271 |
| NYT Connections (extended) | 23.4% | — |
| DTBench | — | 62.9% |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| Epoch Capabilities Index | — | 129 |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math ERNIE 5.1 leads
ERNIE 5.1: 40.3 (#92), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Math | 1481 | 1283 |
| OTIS Mock AIME 2024-2025 | — | 8.1% |
| Omni-MATH | — | 33% |
| MATH Level 5 | — | 63.2% |
Knowledge ERNIE 5.1 leads
ERNIE 5.1: 41.9 (#102), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Expert | 1492 | 1245 |
| GPQA Diamond | — | 49.1% |
| MMLU-Pro | — | 63.1% |
| Confabulations | — | 19.1% |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual ERNIE 5.1 leads
ERNIE 5.1: 55.5 (#29), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1454 | 1252 |
| LMArena Chinese | 1508 | 1272 |
| LMArena French | 1488 | 1280 |
| LMArena German | 1470 | 1234 |
| LMArena Japanese | 1422 | 1180 |
| LMArena Korean | 1427 | 1188 |
| LMArena Russian | 1459 | 1264 |
| LMArena Spanish | 1473 | 1256 |
Instruction Following ERNIE 5.1 leads
ERNIE 5.1: 76.7 (#37), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1460 | 1254 |
| IFEval | — | 80.6% |
Long Context ERNIE 5.1 leads
ERNIE 5.1: 44.7 (#59), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1462 | 1282 |
Writing & Preference ERNIE 5.1 leads
ERNIE 5.1: 65.1 (#52), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | ERNIE 5.1 | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1468 | 1269 |
| LMArena Creative Writing | 1441 | 1221 |
| LMArena Multi-Turn | 1471 | 1272 |
| WildBench | — | 80.2% |
Frequently asked questions
Is ERNIE 5.1 better than Qwen2.5 72B Instruct?
ERNIE 5.1 is the stronger model overall, scoring 43.8 to 31.9 on the Noometry Index.
Is ERNIE 5.1 or Qwen2.5 72B Instruct better for coding?
ERNIE 5.1 scores higher on coding benchmarks: 44.0 versus 33.2 in the Noometry coding category.
How many benchmarks do ERNIE 5.1 and Qwen2.5 72B Instruct share?
17 benchmarks have published results for both models. ERNIE 5.1 has 19 scored results on Noometry and Qwen2.5 72B Instruct has 43.