Model comparison
Nemotron 4 340b Instruct vs Qwen2.5-Max
Qwen2.5-Max is the stronger model overall, scoring 40.7 to 35.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
NVIDIA
Alibaba (Qwen)Summary
- They share 17 benchmarks with published results for both. Nemotron 4 340b Instruct scores higher in 0 categories and Qwen2.5-Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen2.5-Max leads 55.4 to 42.7.
- Nemotron 4 340b Instruct has downloadable open weights; the other is API-only.
Side by side
| Nemotron 4 340b Instruct | Qwen2.5-Max | |
|---|---|---|
| Provider | NVIDIA | Alibaba (Qwen) |
| Noometry Index | 35.9 | 40.7 |
| Released | 2024-06-14 | 2025-01-25 |
| Weights | Open | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 17 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5-Max leads
Nemotron 4 340b Instruct: 35.2 (#228), Qwen2.5-Max: 41.8 (#117)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Coding | 1209 | 1359 |
| LiveBench Coding | — | 64.4% |
Reasoning Qwen2.5-Max leads
Nemotron 4 340b Instruct: 23.4 (#178), Qwen2.5-Max: 25.6 (#147)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Hard Prompts | 1202 | 1360 |
| LiveBench Reasoning | — | 51.4% |
| LiveBench Data Analysis | — | 67.9% |
| Epoch Capabilities Index | — | 132.53 |
| LiveBench | — | 62.3% |
Math Qwen2.5-Max leads
Nemotron 4 340b Instruct: 34.3 (#195), Qwen2.5-Max: 36.9 (#162)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Math | 1216 | 1369 |
| LiveBench Math | — | 58.4% |
Knowledge Qwen2.5-Max leads
Nemotron 4 340b Instruct: 32.0 (#216), Qwen2.5-Max: 35.3 (#186)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Expert | 1171 | 1337 |
| Confabulations | — | 21.8% |
Multilingual Qwen2.5-Max leads
Nemotron 4 340b Instruct: 37.9 (#233), Qwen2.5-Max: 48.1 (#146)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Non-English | 1206 | 1352 |
| LMArena Chinese | 1214 | 1382 |
| LMArena French | 1197 | 1396 |
| LMArena German | 1192 | 1350 |
| LMArena Japanese | 1128 | 1300 |
| LMArena Korean | 1148 | 1304 |
| LMArena Russian | 1219 | 1353 |
| LMArena Spanish | 1195 | 1377 |
Instruction Following Qwen2.5-Max leads
Nemotron 4 340b Instruct: 63.1 (#232), Qwen2.5-Max: 71.3 (#152)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Instruction Following | 1203 | 1335 |
| LiveBench Instruction Following | — | 75.3% |
Long Context Qwen2.5-Max leads
Nemotron 4 340b Instruct: 37.1 (#221), Qwen2.5-Max: 41.4 (#142)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Longer Query | 1225 | 1358 |
Writing & Preference Qwen2.5-Max leads
Nemotron 4 340b Instruct: 42.7 (#233), Qwen2.5-Max: 55.4 (#146)
| Benchmark | Nemotron 4 340b Instruct | Qwen2.5-Max |
|---|---|---|
| LMArena Text | 1225 | 1367 |
| LMArena Creative Writing | 1203 | 1339 |
| LMArena Multi-Turn | 1214 | 1364 |
| Short-Story Creative Writing | — | 72.9% |
| LiveBench Language | — | 56.3% |
Frequently asked questions
Is Nemotron 4 340b Instruct better than Qwen2.5-Max?
Qwen2.5-Max is the stronger model overall, scoring 40.7 to 35.9 on the Noometry Index.
Is Nemotron 4 340b Instruct or Qwen2.5-Max better for coding?
Qwen2.5-Max scores higher on coding benchmarks: 41.8 versus 35.2 in the Noometry coding category.
How many benchmarks do Nemotron 4 340b Instruct and Qwen2.5-Max share?
17 benchmarks have published results for both models. Nemotron 4 340b Instruct has 17 scored results on Noometry and Qwen2.5-Max has 27.