Model comparison
Qwen2.5 72B Instruct vs Qwen3-Next 80B-A3B Instruct
Qwen3-Next 80B-A3B Instruct is the stronger model overall, scoring 43.0 to 31.9 on the Noometry Index.
Last verified . 22 shared benchmarks.
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 1 category and Qwen3-Next 80B-A3B Instruct in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3-Next 80B-A3B Instruct leads 38.8 to 19.3.
- The biggest single-benchmark swing is GPQA (HELM): 42.6% for Qwen2.5 72B Instruct and 63% for Qwen3-Next 80B-A3B Instruct.
- Qwen3-Next 80B-A3B Instruct is cheaper at $0.50 / $2 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
Side by side
| Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 43.0 |
| Released | 2024-09 | 2025-09 |
| Weights | Open | Open |
| Context window | 131K | 131K |
| Max output | 8K | 33K |
| Input $ / M tokens | $1.40 | $0.50 |
| Output $ / M tokens | $5.60 | $2 |
| Results tracked | 43 | 25 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 33.2 (#260), Qwen3-Next 80B-A3B Instruct: 42.5 (#98)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Coding | 1292 | 1440 |
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| BigCodeBench Complete | 55.9% | — |
Agentic & Tool Use Not comparable
Qwen2.5 72B Instruct: 22.1 (#133), Qwen3-Next 80B-A3B Instruct: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
Reasoning Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 22.3 (#199), Qwen3-Next 80B-A3B Instruct: 31.1 (#81)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1271 | 1428 |
| Kagi LLM Benchmark | — | 66.7% |
| 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 Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 19.3 (#287), Qwen3-Next 80B-A3B Instruct: 38.8 (#126)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Omni-MATH | 33% | 46.7% |
| LMArena Math | 1283 | 1440 |
| OTIS Mock AIME 2024-2025 | 8.1% | — |
| MATH Level 5 | 63.2% | — |
Knowledge Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 27.0 (#253), Qwen3-Next 80B-A3B Instruct: 41.8 (#106)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| MMLU-Pro | 63.1% | 78.6% |
| GPQA (HELM) | 42.6% | 63% |
| LMArena Expert | 1245 | 1417 |
| GPQA Diamond | 49.1% | — |
| Confabulations | 19.1% | — |
| Vectara Hallucination Rate | — | 9.3% |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 41.0 (#213), Qwen3-Next 80B-A3B Instruct: 52.1 (#93)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Non-English | 1252 | 1407 |
| LMArena Chinese | 1272 | 1460 |
| LMArena French | 1280 | 1413 |
| LMArena German | 1234 | 1417 |
| LMArena Japanese | 1180 | 1395 |
| LMArena Korean | 1188 | 1364 |
| LMArena Russian | 1264 | 1404 |
| LMArena Spanish | 1256 | 1435 |
Instruction Following Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 65.5 (#221), Qwen3-Next 80B-A3B Instruct: 70.8 (#159)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| IFEval | 80.6% | 81% |
| LMArena Instruction Following | 1254 | 1389 |
Long Context Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 38.9 (#188), Qwen3-Next 80B-A3B Instruct: 37.0 (#223)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Longer Query | 1282 | 1403 |
| Fiction.LiveBench | — | 55.6% |
Writing & Preference Qwen3-Next 80B-A3B Instruct leads
Qwen2.5 72B Instruct: 46.7 (#215), Qwen3-Next 80B-A3B Instruct: 58.0 (#121)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Text | 1269 | 1417 |
| LMArena Creative Writing | 1221 | 1334 |
| WildBench | 80.2% | 80.7% |
| LMArena Multi-Turn | 1272 | 1416 |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen3-Next 80B-A3B Instruct?
Qwen3-Next 80B-A3B Instruct is the stronger model overall, scoring 43.0 to 31.9 on the Noometry Index.
Which is cheaper, Qwen2.5 72B Instruct or Qwen3-Next 80B-A3B Instruct?
Qwen3-Next 80B-A3B Instruct is cheaper. It lists at $0.50 per million input tokens and $2 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Qwen2.5 72B Instruct or Qwen3-Next 80B-A3B Instruct better for coding?
Qwen3-Next 80B-A3B Instruct scores higher on coding benchmarks: 42.5 versus 33.2 in the Noometry coding category.
Which has the bigger context window?
Both accept 131K tokens.
How many benchmarks do Qwen2.5 72B Instruct and Qwen3-Next 80B-A3B Instruct share?
22 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3-Next 80B-A3B Instruct has 25.