Model comparison
Qwen2.5 72B Instruct vs Qwen2.5 7B Instruct
Qwen2.5 72B Instruct is the stronger model overall, scoring 31.5 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 8.0× less per token, which makes it the better buy when Qwen2.5 72B Instruct's lead doesn't matter for your workload.
Last verified . 14 shared benchmarks.
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 4 categories and Qwen2.5 7B Instruct in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen2.5 72B Instruct leads 26.9 to 17.0.
- The biggest single-benchmark swing is DTBench: 62.9% for Qwen2.5 72B Instruct and 47.7% for Qwen2.5 7B Instruct.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $1.40 / $5.60 for Qwen2.5 72B Instruct.
Side by side
| Qwen2.5 72B Instruct | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.5 | 28.6 |
| Released | 2024-09 | 2024-09 |
| Weights | Open | Open |
| Context window | 131K | 131K |
| Max output | 8K | 8K |
| Input $ / M tokens | $1.40 | $0.17 |
| Output $ / M tokens | $5.60 | $0.70 |
| Results tracked | 43 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5 7B Instruct leads
Qwen2.5 72B Instruct: 32.8 (#256), Qwen2.5 7B Instruct: 35.9 (#204)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| BigCodeBench Instruct | 45.8% | 37.6% |
| BigCodeBench Complete | 55.9% | 46.1% |
| WeirdML | 16% | — |
| LMArena Coding | 1293 | — |
Agentic & Tool Use Qwen2.5 7B Instruct leads
Qwen2.5 72B Instruct: 22.0 (#124), Qwen2.5 7B Instruct: 23.5 (#117)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| BALROG | 16.2% | 7.8% |
| TheAgentCompany | 5.7% | — |
| METR Time Horizons | 35.8% | — |
Reasoning Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 21.4 (#197), Qwen2.5 7B Instruct: 14.3 (#317)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| DTBench | 62.9% | 47.7% |
| LMCA | 13.4% | 6.4% |
| Epoch Capabilities Index | 129.01 | 118.44 |
| Chess Puzzles | — | 0% |
| LMArena Hard Prompts | 1271 | — |
| BIG-Bench Hard | 79.8% | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 18.1 (#281), Qwen2.5 7B Instruct: 11.8 (#300)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 2.5% |
| Omni-MATH | 33% | 29.4% |
| LMArena Math | 1283 | — |
| MATH Level 5 | 63.2% | — |
Knowledge Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 26.9 (#247), Qwen2.5 7B Instruct: 17.0 (#281)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 49.1% | 35.5% |
| MMLU-Pro | 63.1% | 53.9% |
| GPQA (HELM) | 42.6% | 34.1% |
| MMLU | 85.3% | 72.9% |
| Confabulations | 19.1% | — |
| LMArena Expert | 1245 | — |
| ARC (AI2) Challenge | 94.5% | — |
| TriviaQA | 71.9% | — |
Multilingual Not comparable
Qwen2.5 72B Instruct: 41.2 (#208), Qwen2.5 7B Instruct: —
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1252 | — |
| LMArena Chinese | 1272 | — |
| LMArena French | 1279 | — |
| LMArena German | 1234 | — |
| LMArena Japanese | 1180 | — |
| LMArena Korean | 1188 | — |
| LMArena Russian | 1264 | — |
| LMArena Spanish | 1255 | — |
Instruction Following Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 65.7 (#216), Qwen2.5 7B Instruct: 63.4 (#226)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | 80.6% | 74.1% |
| LMArena Instruction Following | 1254 | — |
Long Context Not comparable
Qwen2.5 72B Instruct: 38.9 (#183), Qwen2.5 7B Instruct: —
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1282 | — |
Writing & Preference Qwen2.5 7B Instruct leads
Qwen2.5 72B Instruct: 46.9 (#210), Qwen2.5 7B Instruct: 48.8 (#194)
| Benchmark | Qwen2.5 72B Instruct | Qwen2.5 7B Instruct |
|---|---|---|
| WildBench | 80.2% | 73.1% |
| LMArena Text | 1269 | — |
| LMArena Creative Writing | 1221 | — |
| LMArena Multi-Turn | 1272 | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen2.5 7B Instruct?
Qwen2.5 72B Instruct is the stronger model overall, scoring 31.5 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 8.0× less per token, which makes it the better buy when Qwen2.5 72B Instruct's lead doesn't matter for your workload.
Which is cheaper, Qwen2.5 72B Instruct or Qwen2.5 7B Instruct?
Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; Qwen2.5 72B Instruct lists at $1.40 and $5.60.
Is Qwen2.5 72B Instruct or Qwen2.5 7B Instruct better for coding?
Qwen2.5 7B Instruct scores higher on coding benchmarks: 35.9 versus 32.8 in the Noometry coding category.
Which has the bigger context window?
Both accept 131K tokens.
How many benchmarks do Qwen2.5 72B Instruct and Qwen2.5 7B Instruct share?
14 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen2.5 7B Instruct has 15.