Grok 4.20 vs Qwen3 8
Grok 4.20 
Qwen3 8 Verdict
- Qwen3 8 (max) leads on quality: 66.5 vs 59.9.
- Grok 4.20 (thinking) is stronger in instruction following.
- Qwen3 8 (max) is stronger in agents & tools, coding, composite, human preference, knowledge, long context, maths, multimodal, reasoning.
- Grok 4.20 (thinking) is 1.7× cheaper ($1.56 vs $2.67 per 1M blended).
- Grok 4.20 (thinking) streams 2.6× faster (101 vs 39 tokens per second).
| Metric | Grok 4.20 (thinking) | Qwen3 8 (max) |
|---|---|---|
| BenchLeader Index | 59.9 | 66.5 |
| Agents & tools score | 51.7 | 68.3 |
| Coding score | 54.7 | 66.5 |
| Composite score | 61.5 | 74.3 |
| Human preference score | 67.1 | 68.2 |
| Instruction following score | 79.9 | – |
| Knowledge score | 56.8 | 63.5 |
| Long context score | 60.7 | 66.6 |
| Maths score | 55.6 | 67.0 |
| Multimodal score | 59.9 | 67.4 |
| Reasoning score | 60.7 | 76.3 |
| Blended price $/M | $1.56 | $2.67 |
| Output speed | 101 tok/s | 39 tok/s |
| Time to first answer | 21.4 s | 55.1 s |
| Context window | 1M | 1M |
| GPQA Diamond | 89.3% | – |
| FrontierMath Tiers 1–3 | 44.9% | – |
| FrontierMath Tier 4 | 17.1% | – |
| OTIS Mock AIME | 92.2% | – |
| SimpleQA Verified | 30.2% | – |
| SciCode | – | 52.9% |
| ProofBench | 14.0% | 58.0% |
| Epoch Capabilities Index | – | 156.6 |
| LMArena Text | 1472 | 1481 |
| LMArena Hard Prompts | 1488 | 1503 |
| LMArena Coding | 1510 | 1522 |
| LMArena WebDev | 1374 | 1671 |
| LMArena Vision | 1263 | 1315 |
| LMArena Agent | – | 3.3 |
| LiveBench | – | 78.5% |
| LiveBench Reasoning | – | 88.2% |
| LiveBench Coding | – | 72.9% |
| LiveBench Agentic Coding | – | 64.7% |
| LiveBench Mathematics | – | 91.3% |
| LiveBench Data Analysis | – | 78.4% |
| LiveBench Language | – | 79.7% |
| LiveBench Instruction Following | – | 74.1% |
| AA Intelligence Index | 25.7 | 45.4 |
| IFBench | 82.9% | – |
| AA-LCR | 69.0% | 80.3% |
| MMMU-Pro | 74.6% | 82.8% |
| AA-Omniscience | 14.8 | 12.0 |
| Terminal-Bench Hard | 40.9% | – |
| GPQA Diamond (AA) | 91.1% | 92.8% |
| Humanity's Last Exam (AA) | 34.5% | 43.1% |
| SciCode (AA) | – | 53.2% |
| τ²-Bench Telecom (AA) | 96.5% | – |
| AIME (Vals) | 96.5% | – |
| LiveCodeBench | 84.3% | 87.8% |
| MMLU-Pro | 86.3% | 88.6% |
| IOI | – | 68.9% |
| LegalBench | 77.7% | 83.6% |
| CorpFin | 63.7% | 65.8% |
| TaxEval | 74.1% | 75.5% |
| MedQA | 94.5% | – |
| Terminal-Bench 2.1 (Vals) | 44.2% | 67.4% |
| SWE-bench (Vals) | 72.2% | 85.6% |
| GPQA Diamond (Vals) | 88.6% | 93.7% |
| Vals Index | 17.6 | 51.8 |
| Kagi LLM Benchmark | 75.0% | – |
| ARC-AGI-1 | 89.5% | – |
| ARC-AGI-2 | 65.1% | – |
| ARC-AGI-3 | 0.1% | – |
| CritPt | 6.6% | 20.0% |
| GDPval (AA) | – | 58.2% |
| τ²-Bench Banking (AA) | – | 51.3% |
| APEX-Agents (AA) | 14.2% | – |
| CaseLaw v2 | 54.5% | – |
| Code Migration | 0.3% | 24.0% |
| Excel Modeling Benchmark | 11.7% | 60.1% |
| Finance Agent v2 | 28.5% | 50.6% |
| Harvey's Legal Agent Benchmark | 0.0% | 10.4% |
| Legal Research Bench | 13.9% | 47.6% |
| MedCode | 32.2% | 40.7% |
| MedScribe | 63.4% | 85.0% |
| MMMU-Pro (Vals) | 83.5% | 88.0% |
| MortgageTax | 45.4% | 64.0% |
| MysteryMechanism | – | 23.9% |
| ProgramBench | – | 0.0% |
| Public Benefits Bench | – | 67.1% |
| SAGE | 38.2% | 51.3% |
| SkillsBench | – | 42.0% |
| Tax Agent Bench | – | 66.0% |
| Terminal-Bench 2.0 (Vals) | 40.5% | – |
| Terminal-Bench 4.0 (Vals) | – | 24.8% |
| Terminal-Bench Science | – | 1.4% |
| Vals Multimodal Index | 39.1% | 65.4% |
| Vibe Code Bench 1-100 | – | 12.8% |
| Vibe Code Bench v1.1 | 4.1% | 64.7% |
| LMArena Maths | 1467 | 1498 |
| LMArena Creative Writing | 1446 | 1472 |
| LMArena Instruction Following | 1445 | 1477 |
| LMArena Multi-turn | 1480 | 1495 |
| LMArena Longer Queries | 1465 | 1492 |
| LMArena Document | 1439 | – |
| Chess Puzzles | 24.0% | – |
| ForecastBench | 60.7% | – |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Grok 4.20 vs Qwen3 8: questions
- Is Grok 4.20 better than Qwen3 8?
- Qwen3 8 (max) leads on quality: 66.5 vs 59.9. The BenchLeader Index combines every independent quality benchmark; Qwen3 8 (max) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Grok 4.20 better than Qwen3 8 for coding?
- Qwen3 8 scores higher in coding (67 vs 55 on the category index, where 50 is average).
- Is Grok 4.20 better than Qwen3 8 for agentic tasks?
- Qwen3 8 scores higher in agentic tasks (68 vs 52 on the category index, where 50 is average).
- Which is cheaper, Grok 4.20 or Qwen3 8?
- Grok 4.20 is cheaper: $1.56 against $2.67 per million tokens, blended at three input tokens per output token.
- Which is faster, Grok 4.20 or Qwen3 8?
- Grok 4.20 streams faster: 101 against 39 output tokens per second.
- Which has the larger context window?
- Both accept 1M tokens of context.