Model comparison
Grok 4 vs Qwen3-Next 80B-A3B Instruct
Grok 4 is the stronger model overall, scoring 47.8 to 42.5 on the Noometry Index.
Last verified . 24 shared benchmarks.
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Grok 4 scores higher in 7 categories and Qwen3-Next 80B-A3B Instruct in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in long context, where Grok 4 leads 63.1 to 37.0.
- The biggest single-benchmark swing is Fiction.LiveBench: 94.4% for Grok 4 and 55.6% for Qwen3-Next 80B-A3B Instruct.
- Qwen3-Next 80B-A3B Instruct has downloadable open weights; the other is API-only.
Side by side
| Grok 4 | Qwen3-Next 80B-A3B Instruct | |
|---|---|---|
| Provider | xAI | Alibaba (Qwen) |
| Noometry Index | 47.8 | 42.5 |
| Released | 2025-07-09 | 2025-09 |
| Weights | Proprietary | Open |
| Context window | — | 131K |
| Max output | — | 33K |
| Input $ / M tokens | — | $0.50 |
| Output $ / M tokens | — | $2 |
| Results tracked | 48 | 25 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
Grok 4: 49.9 (#44), Qwen3-Next 80B-A3B Instruct: 41.9 (#101)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Coding | 1408 | 1441 |
| Aider Polyglot | 79.6% | — |
| WeirdML | 45.7% | — |
Agentic & Tool Use Not comparable
Grok 4: 32.1 (#62), Qwen3-Next 80B-A3B Instruct: —
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Terminal-Bench | 27.2% | — |
| Berkeley Function Calling Leaderboard | 63% | — |
| GDPval | 21.1% | — |
| Cybench | 43% | — |
| DeepResearch Bench | 47.3% | — |
| BALROG | 43.6% | — |
| LMArena Search | 1142 | — |
| METR Time Horizons | 66.6% | — |
Reasoning Grok 4 leads
Grok 4: 36.9 (#59), Qwen3-Next 80B-A3B Instruct: 30.1 (#79)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 73.6% | 66.7% |
| LMArena Hard Prompts | 1409 | 1428 |
| ARC-AGI-2 | 16% | — |
| SimpleBench | 60.5% | — |
| ARC-AGI-1 | 66.7% | — |
| Chess Puzzles | 28% | — |
| Epoch Capabilities Index | 146.46 | — |
| ForecastBench | 60.9 | — |
Math Grok 4 leads
Grok 4: 46.5 (#58), Qwen3-Next 80B-A3B Instruct: 36.3 (#118)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Omni-MATH | 60.3% | 46.7% |
| LMArena Math | 1422 | 1440 |
| OTIS Mock AIME 2024-2025 | 84% | — |
| FrontierMath (Feb 2025 set) | 19.7% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Grok 4 leads
Grok 4: 53.7 (#50), Qwen3-Next 80B-A3B Instruct: 41.3 (#102)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| MMLU-Pro | 85.1% | 78.6% |
| GPQA (HELM) | 72.7% | 63% |
| LMArena Expert | 1417 | 1416 |
| GPQA Diamond | 87% | — |
| Confabulations | 12.4% | — |
| Vectara Hallucination Rate | — | 9.3% |
Multimodal Not comparable
Grok 4: 33.1 (#89), Qwen3-Next 80B-A3B Instruct: —
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Vision | 1211 | — |
| GeoBench | 45% | — |
Multilingual Too close to call
Grok 4: 52.0 (#98), Qwen3-Next 80B-A3B Instruct: 52.3 (#88)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Non-English | 1403 | 1407 |
| LMArena Chinese | 1428 | 1461 |
| LMArena French | 1418 | 1415 |
| LMArena German | 1429 | 1416 |
| LMArena Japanese | 1396 | 1398 |
| LMArena Korean | 1376 | 1366 |
| LMArena Russian | 1409 | 1403 |
| LMArena Spanish | 1419 | 1436 |
Instruction Following Grok 4 leads
Grok 4: 79.4 (#4), Qwen3-Next 80B-A3B Instruct: 71.0 (#154)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| IFEval | 94.9% | 81% |
| LMArena Instruction Following | 1387 | 1388 |
Long Context Grok 4 leads
Grok 4: 63.1 (#4), Qwen3-Next 80B-A3B Instruct: 37.0 (#218)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Fiction.LiveBench | 94.4% | 55.6% |
| LMArena Longer Query | 1409 | 1403 |
Writing & Preference Too close to call
Grok 4: 58.7 (#111), Qwen3-Next 80B-A3B Instruct: 58.2 (#116)
| Benchmark | Grok 4 | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Text | 1412 | 1417 |
| LMArena Creative Writing | 1397 | 1333 |
| WildBench | 79.7% | 80.7% |
| LMArena Multi-Turn | 1416 | 1415 |
| Short-Story Creative Writing | 76.9% | — |
Frequently asked questions
Is Grok 4 better than Qwen3-Next 80B-A3B Instruct?
Grok 4 is the stronger model overall, scoring 47.8 to 42.5 on the Noometry Index.
Is Grok 4 or Qwen3-Next 80B-A3B Instruct better for coding?
Grok 4 scores higher on coding benchmarks: 49.9 versus 41.9 in the Noometry coding category.
How many benchmarks do Grok 4 and Qwen3-Next 80B-A3B Instruct share?
24 benchmarks have published results for both models. Grok 4 has 48 scored results on Noometry and Qwen3-Next 80B-A3B Instruct has 25.