Model comparison
Nemotron 4 340b Instruct vs Qwen3.5 27B
Qwen3.5 27B is the stronger model overall, scoring 41.9 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 Qwen3.5 27B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen3.5 27B leads 59.3 to 42.7.
Side by side
| Nemotron 4 340b Instruct | Qwen3.5 27B | |
|---|---|---|
| Provider | NVIDIA | Alibaba (Qwen) |
| Noometry Index | 35.9 | 41.9 |
| Released | 2024-06-14 | 2026-02-23 |
| Weights | Open | Open |
| Context window | — | 262K |
| Max output | — | 66K |
| Input $ / M tokens | — | $0.30 |
| Output $ / M tokens | — | $2.40 |
| Results tracked | 17 | 28 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 27B leads
Nemotron 4 340b Instruct: 35.2 (#228), Qwen3.5 27B: 38.9 (#168)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Coding | 1209 | 1427 |
| LMArena WebDev | — | 1358 |
| WeirdML | — | 39.5% |
| ALE-Bench | — | 349.45 |
Agentic & Tool Use Not comparable
Nemotron 4 340b Instruct: —, Qwen3.5 27B: —
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| Vending-Bench 2 | — | 201.98 |
Reasoning Qwen3.5 27B leads
Nemotron 4 340b Instruct: 23.4 (#178), Qwen3.5 27B: 27.5 (#117)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Hard Prompts | 1202 | 1414 |
| NYT Connections (extended) | — | 47.9% |
| Thematic Generalization | — | 45.5% |
| DTBench | — | 82.4% |
| LMCA | — | 34% |
Math Qwen3.5 27B leads
Nemotron 4 340b Instruct: 34.3 (#195), Qwen3.5 27B: 38.8 (#127)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Math | 1216 | 1429 |
| MathArena Final-Answer Competitions | — | 56.7% |
Knowledge Qwen3.5 27B leads
Nemotron 4 340b Instruct: 32.0 (#216), Qwen3.5 27B: 38.0 (#150)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Expert | 1171 | 1428 |
| Vectara Hallucination Rate | — | 12.1% |
Multimodal Not comparable
Nemotron 4 340b Instruct: —, Qwen3.5 27B: 39.4 (#59)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Vision | — | 1241 |
Multilingual Qwen3.5 27B leads
Nemotron 4 340b Instruct: 37.9 (#233), Qwen3.5 27B: 50.8 (#115)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Non-English | 1206 | 1390 |
| LMArena Chinese | 1214 | 1478 |
| LMArena French | 1197 | 1410 |
| LMArena German | 1192 | 1393 |
| LMArena Japanese | 1128 | 1345 |
| LMArena Korean | 1148 | 1358 |
| LMArena Russian | 1219 | 1390 |
| LMArena Spanish | 1195 | 1407 |
Instruction Following Qwen3.5 27B leads
Nemotron 4 340b Instruct: 63.1 (#232), Qwen3.5 27B: 73.5 (#119)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Instruction Following | 1203 | 1393 |
Long Context Qwen3.5 27B leads
Nemotron 4 340b Instruct: 37.1 (#221), Qwen3.5 27B: 43.1 (#106)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Longer Query | 1225 | 1413 |
Writing & Preference Qwen3.5 27B leads
Nemotron 4 340b Instruct: 42.7 (#233), Qwen3.5 27B: 59.3 (#111)
| Benchmark | Nemotron 4 340b Instruct | Qwen3.5 27B |
|---|---|---|
| LMArena Text | 1225 | 1409 |
| LMArena Creative Writing | 1203 | 1362 |
| LMArena Multi-Turn | 1214 | 1410 |
Frequently asked questions
Is Nemotron 4 340b Instruct better than Qwen3.5 27B?
Qwen3.5 27B is the stronger model overall, scoring 41.9 to 35.9 on the Noometry Index.
Is Nemotron 4 340b Instruct or Qwen3.5 27B better for coding?
Qwen3.5 27B scores higher on coding benchmarks: 38.9 versus 35.2 in the Noometry coding category.
How many benchmarks do Nemotron 4 340b Instruct and Qwen3.5 27B share?
17 benchmarks have published results for both models. Nemotron 4 340b Instruct has 17 scored results on Noometry and Qwen3.5 27B has 28.