Model comparison
Grok-2 (Dec 2024) vs Qwen2.5 72B Instruct
Grok-2 (Dec 2024) is the stronger model overall, scoring 33.7 to 31.9 on the Noometry Index.
Last verified . 24 shared benchmarks.
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Grok-2 (Dec 2024) scores higher in 6 categories and Qwen2.5 72B Instruct in 2 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen2.5 72B Instruct leads 22.3 to 16.9.
- The biggest single-benchmark swing is WeirdML: 22.2% for Grok-2 (Dec 2024) and 16% for Qwen2.5 72B Instruct.
- Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| Grok-2 (Dec 2024) | Qwen2.5 72B Instruct | |
|---|---|---|
| Provider | xAI | Alibaba (Qwen) |
| Noometry Index | 33.7 | 31.9 |
| Released | 2024-08-13 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 8K |
| Input $ / M tokens | — | $1.40 |
| Output $ / M tokens | — | $5.60 |
| Results tracked | 34 | 43 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Grok-2 (Dec 2024): 33.3 (#258), Qwen2.5 72B Instruct: 33.2 (#260)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| WeirdML | 22.2% | 16% |
| LMArena Coding | 1287 | 1292 |
| BigCodeBench Instruct | — | 45.8% |
| LiveBench Coding | 46.4% | — |
| BigCodeBench Complete | — | 55.9% |
Agentic & Tool Use Not comparable
Grok-2 (Dec 2024): —, Qwen2.5 72B Instruct: 22.1 (#133)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| TheAgentCompany | — | 5.7% |
| BALROG | — | 16.2% |
| METR Time Horizons | — | 35.8% |
Reasoning Qwen2.5 72B Instruct leads
Grok-2 (Dec 2024): 16.9 (#299), Qwen2.5 72B Instruct: 22.3 (#199)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Hard Prompts | 1272 | 1271 |
| DTBench | 65.2% | 62.9% |
| Epoch Capabilities Index | 130.48 | 129 |
| SimpleBench | 22.7% | — |
| LiveBench Reasoning | 54.8% | — |
| LiveBench Data Analysis | 54.5% | — |
| LMCA | — | 13.4% |
| BIG-Bench Hard | — | 79.8% |
| ForecastBench | — | 57.5 |
| HellaSwag | — | 84.8% |
| LiveBench | 54.3% | — |
| PIQA | — | 82.6% |
| WinoGrande | — | 82.3% |
Math Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 20.8 (#284), Qwen2.5 72B Instruct: 19.3 (#287)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 11.5% | 8.1% |
| LMArena Math | 1283 | 1283 |
| MATH Level 5 | 63.5% | 63.2% |
| Omni-MATH | — | 33% |
| LiveBench Math | 54.9% | — |
| FrontierMath (Feb 2025 set) | 0.7% | — |
Knowledge Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 29.8 (#233), Qwen2.5 72B Instruct: 27.0 (#253)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| GPQA Diamond | 53.8% | 49.1% |
| Confabulations | 20.1% | 19.1% |
| LMArena Expert | 1254 | 1245 |
| MMLU-Pro | — | 63.1% |
| GPQA (HELM) | — | 42.6% |
| ARC (AI2) Challenge | — | 94.5% |
| MMLU | — | 85.3% |
| TriviaQA | — | 71.9% |
Multilingual Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 43.1 (#188), Qwen2.5 72B Instruct: 41.0 (#213)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Non-English | 1282 | 1252 |
| LMArena Chinese | 1289 | 1272 |
| LMArena French | 1318 | 1280 |
| LMArena German | 1287 | 1234 |
| LMArena Japanese | 1244 | 1180 |
| LMArena Korean | 1237 | 1188 |
| LMArena Russian | 1286 | 1264 |
| LMArena Spanish | 1281 | 1256 |
Instruction Following Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 66.9 (#202), Qwen2.5 72B Instruct: 65.5 (#221)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Instruction Following | 1270 | 1254 |
| LiveBench Instruction Following | 69.6% | — |
| IFEval | — | 80.6% |
Long Context Too close to call
Grok-2 (Dec 2024): 38.8 (#190), Qwen2.5 72B Instruct: 38.9 (#188)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Longer Query | 1276 | 1282 |
Writing & Preference Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 48.6 (#198), Qwen2.5 72B Instruct: 46.7 (#215)
| Benchmark | Grok-2 (Dec 2024) | Qwen2.5 72B Instruct |
|---|---|---|
| LMArena Text | 1305 | 1269 |
| LMArena Creative Writing | 1284 | 1221 |
| LMArena Multi-Turn | 1290 | 1272 |
| Short-Story Creative Writing | 63.6% | — |
| WildBench | — | 80.2% |
| LiveBench Language | 45.6% | — |
Frequently asked questions
Is Grok-2 (Dec 2024) better than Qwen2.5 72B Instruct?
Grok-2 (Dec 2024) is the stronger model overall, scoring 33.7 to 31.9 on the Noometry Index.
Is Grok-2 (Dec 2024) or Qwen2.5 72B Instruct better for coding?
They score almost the same on coding (33.3 vs 33.2); test both on your own repository before choosing.
How many benchmarks do Grok-2 (Dec 2024) and Qwen2.5 72B Instruct share?
24 benchmarks have published results for both models. Grok-2 (Dec 2024) has 34 scored results on Noometry and Qwen2.5 72B Instruct has 43.