Grok 4.5 vs Qwen3.8 Max (0902)
Updated October 21, 2026. 105 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Grok 4.5 for coding: it scores 75.5 on the Vibe Coding Index against 63.5 for Qwen3.8 Max (0902).
Pick Qwen3.8 Max (0902) when the context window is larger, 1M against 500K tokens, or when what matters is LMArena WebDev or AA Coding Index, where it leads.
- Shared results
- 105
- Grok 4.5 only
- 15
- Qwen3.8 Max (0902) only
- 8
- Like-for-like categories
- 8 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, four public coding benchmarks
Grok 4.5
Grok 4.5 scores 75.5 against 63.5.
2 benchmarks behind each score
Intelligence
AA Intelligence Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 45.4 against 38.8.
Same board, both measured
Agentic work
AA Agentic Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 56.0 against 41.2.
Same board, both measured
Reasoning
GPQA Diamond
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 93.7% against 92.9%.
Same board, both measured
Knowledge
MMLU Pro
Grok 4.5
Grok 4.5 leads, 89.2% against 88.6%.
Same board, both measured
Web design
Design Arena: websites
Grok 4.5
Grok 4.5 leads, 1281 against 1275.
Same board, both measured
Long documents
Listed context window
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) takes 1M tokens against 500K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
No call
Grok 4.5 $0.0050, Qwen3.8 Max (0902) $0.0050 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
Grok 4.5
Grok 4.5 $0.62, Qwen3.8 Max (0902) $1.32.
Published attempt cost and pass rate
At a glance
- Coding index
- Grok 4.5 75.5 from 2 benchmarks, Qwen3.8 Max (0902) 63.5 from 2 benchmarks.
- Widest category
- Math, 12.9 points apart on average over 6 shared benchmarks.
- Widest single gap
- CyberBench (Vals AI): Grok 4.5 71.0% against Qwen3.8 Max (0902) 28.6%, 42.4 points apart on a 0 to 100 scale.
- Output price
- Both cost $6.00 per million output tokens.
- Context window
- Grok 4.5 takes 500K tokens, Qwen3.8 Max (0902) 1M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Grok 4.5 | Bars | Qwen3.8 Max (0902) | Ties |
|---|---|---|---|---|
| Coding | 8 | 10 | 1 tie | |
| Intelligence | 0 | 2 | ||
| Agentic work | 3 | 13 | ||
| Reasoning | 2 | 10 | ||
| Math | 0 | 6 | ||
| Knowledge | 5 | 6 | ||
| Web design | 2 | 6 | ||
| General | 3 | 28 | ||
| All shared boards | 23 | 81 | 1 tie |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index810
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| LiveBench coding | 94.3% | 60.3% | Grok 4.5 |
| LMArena WebDev | 1553 | 1674 | Qwen3.8 Max (0902) |
| AA Coding Index | 72.4 | 76.2 | Qwen3.8 Max (0902) |
| LiveCodeBench (Vals AI) | 87.3% | 87.8% | Qwen3.8 Max (0902) |
| SWE-bench Verified (Vals AI) | 86.6% | 85.6% | Grok 4.5 |
| Terminal-Bench 4 (Vals AI) | 8.6% | 34.3% | Qwen3.8 Max (0902) |
| Vibe Code Bench (Vals AI) | 69.0% | 64.7% | Grok 4.5 |
| Code migration (Vals AI) | 36.6% | 24.0% | Grok 4.5 |
| ProgramBench (Vals AI) | 0.0% | 0.0% | Tie |
| IOI (Vals AI) | 40.6% | 68.9% | Qwen3.8 Max (0902) |
| Terminal-Bench 2.1 (Vals AI) | 67.8% | 67.4% | Grok 4.5 |
| SWE-bench Multimodal (Vals AI) | Not measured | 42.5% | Only Qwen3.8 Max (0902) measured |
| Vibe Code Bench 1-100 (Vals AI) | Not measured | 12.8% | Only Qwen3.8 Max (0902) measured |
| SkillsBench (Vals AI) | 66.0% | 42.0% | Grok 4.5 |
| CyberBench (Vals AI) | 71.0% | 28.6% | Grok 4.5 |
| SciCode | 55.0% | 53.2% | Grok 4.5 |
| FrontierCode | 42.4% | Not measured | Only Grok 4.5 measured |
| FrontierSWE | Not measured | 17.8% | Only Qwen3.8 Max (0902) measured |
| WeirdML | 46.4% | Not measured | Only Grok 4.5 measured |
| ALE-Bench | 1309.3 | Not measured | Only Grok 4.5 measured |
| LMArena Text: coding | 1474 | 1502 | Qwen3.8 Max (0902) |
| LMArena Text: software and IT | 1471 | 1498 | Qwen3.8 Max (0902) |
| LMArena WebDev: React | 1556 | 1692 | Qwen3.8 Max (0902) |
| LMArena WebDev: HTML | 1567 | 1649 | Qwen3.8 Max (0902) |
| LMArena WebDev: from an image | 1562 | 1637 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 13 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| AA Intelligence Index | 38.8 | 45.4 | Qwen3.8 Max (0902) |
| Epoch Capabilities Index | 154.1 | 157.4 | Qwen3.8 Max (0902) |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index313
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| AA Agentic Index | 41.2 | 56.0 | Qwen3.8 Max (0902) |
| Design Arena: agents building web apps | 1217 | 1301 | Qwen3.8 Max (0902) |
| Design Arena: agents building full-stack apps | 1249 | 1305 | Qwen3.8 Max (0902) |
| Design Arena: agents building mobile apps | 1221 | 1223 | Qwen3.8 Max (0902) |
| Design Arena: agents building Android apps | 1242 | 1207 | Grok 4.5 |
| Design Arena: agents building Godot games | 1229 | 1234 | Qwen3.8 Max (0902) |
| Design Arena: agents building games | 1235 | Not measured | Only Grok 4.5 measured |
| Design Arena: HTML slides | 1204 | Not measured | Only Grok 4.5 measured |
| Design Arena: PowerPoint slides | 1232 | 1257 | Qwen3.8 Max (0902) |
| Design Arena: spreadsheets | Not measured | 868 | Only Qwen3.8 Max (0902) measured |
| LiveBench research | 85.9% | 69.9% | Grok 4.5 |
| LiveBench browser | 44.0% | 32.0% | Grok 4.5 |
| LiveBench long agentic | 27.9% | 42.3% | Qwen3.8 Max (0902) |
| SREBench (Vals AI) | 0.8% | Not measured | Only Grok 4.5 measured |
| Finance Agent (Vals AI) | 48.4% | 50.6% | Qwen3.8 Max (0902) |
| Tax Agent Bench (Vals AI) | 24.9% | 32.1% | Qwen3.8 Max (0902) |
| Legal Research Bench (Vals AI) | 38.0% | 47.6% | Qwen3.8 Max (0902) |
| Excel Modeling Benchmark (Vals AI) | 52.9% | 60.1% | Qwen3.8 Max (0902) |
| Terminal-Bench-Science (Vals AI) | Not measured | 12.9% | Only Qwen3.8 Max (0902) measured |
| Time Horizon Index (Vals AI) | 6.3% | Not measured | Only Grok 4.5 measured |
| APEX-Agents | 56.2% | 63.3% | Qwen3.8 Max (0902) |
| PostTrainBench | 23.4% | Not measured | Only Grok 4.5 measured |
| Vending-Bench 2 | $3,887 | Not measured | Only Grok 4.5 measured |
| LMArena Search | 1216 | 1261 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 15 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index210
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| GPQA Diamond | 92.9% | 93.7% | Qwen3.8 Max (0902) |
| LiveBench reasoning | 87.2% | 88.2% | Qwen3.8 Max (0902) |
| MysteryMechanism (Vals AI) | Not measured | 23.9% | Only Qwen3.8 Max (0902) measured |
| MMMU Pro (Vals AI) | 61.8% | 88.0% | Qwen3.8 Max (0902) |
| GPQA Diamond (Artificial Analysis run) | 93.1% | 92.8% | Grok 4.5 |
| Humanity's Last Exam (Artificial Analysis run) | 42.7% | 43.1% | Qwen3.8 Max (0902) |
| GPQA Diamond (Epoch AI run) | 93.4% | 92.7% | Grok 4.5 |
| ARC-AGI | 87.2% | Not measured | Only Grok 4.5 measured |
| ARC-AGI-2 | 52.6% | Not measured | Only Grok 4.5 measured |
| SimpleBench | 70.0% | Not measured | Only Grok 4.5 measured |
| Chess puzzles | 35.0% | 46.0% | Qwen3.8 Max (0902) |
| Mystery game puzzles | Not measured | 38.0% | Only Qwen3.8 Max (0902) measured |
| LMArena Text: hard prompts | 1462 | 1496 | Qwen3.8 Max (0902) |
| LMArena Text: expert | 1465 | 1507 | Qwen3.8 Max (0902) |
| LMArena Text: hard prompts in English | 1464 | 1503 | Qwen3.8 Max (0902) |
| LMArena Vision: diagrams | 1304 | 1323 | Qwen3.8 Max (0902) |
| LMArena Vision: homework | 1295 | 1336 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| OTIS Mock AIME | 97.8% | 100.0% | Qwen3.8 Max (0902) |
| FrontierMath, tiers 1 to 3 | 57.2% | 74.7% | Qwen3.8 Max (0902) |
| FrontierMath, tier 4 | 24.4% | 46.3% | Qwen3.8 Max (0902) |
| ProofBench | 31.0% | 58.0% | Qwen3.8 Max (0902) |
| LMArena Text: math | 1460 | 1501 | Qwen3.8 Max (0902) |
| LMArena Text: mathematical work | 1482 | 1491 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index56
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| MMLU Pro | 89.2% | 88.6% | Grok 4.5 |
| LegalBench (Vals AI) | 86.0% | 83.6% | Grok 4.5 |
| CorpFin (Vals AI) | 67.4% | 65.8% | Grok 4.5 |
| TaxEval (Vals AI) | 71.7% | 75.5% | Qwen3.8 Max (0902) |
| MortgageTax (Vals AI) | 61.8% | 64.0% | Qwen3.8 Max (0902) |
| MedCode (Vals AI) | 43.3% | 40.7% | Grok 4.5 |
| SimpleQA Verified | 48.3% | 47.3% | Grok 4.5 |
| LMArena Text: science | 1464 | 1512 | Qwen3.8 Max (0902) |
| LMArena Text: medicine and healthcare | 1440 | 1503 | Qwen3.8 Max (0902) |
| LMArena Text: legal and government | 1463 | 1492 | Qwen3.8 Max (0902) |
| LMArena Text: business and finance | 1441 | 1480 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 6 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index26
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| Design Arena: websites | 1281 | 1275 | Grok 4.5 |
| Design Arena: UI components | 1279 | 1340 | Qwen3.8 Max (0902) |
| Design Arena: game dev | 1263 | 1313 | Qwen3.8 Max (0902) |
| Design Arena: data viz | 1267 | 1281 | Qwen3.8 Max (0902) |
| Design Arena: 3D | 1257 | 1337 | Qwen3.8 Max (0902) |
| Design Arena: all code categories | 1276 | 1293 | Qwen3.8 Max (0902) |
| Design Arena: SVG | 1215 | Not measured | Only Grok 4.5 measured |
| Design Arena: ASCII art | 1257 | 1201 | Grok 4.5 |
| LiveBench design | 78.1% | 78.6% | Qwen3.8 Max (0902) |
GeneralLike-for-likePublic boards, shown as published, kept out of the index328
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| LiveBench data analysis | 73.0% | 78.4% | Qwen3.8 Max (0902) |
| Public Benefits Bench (Vals AI) | Not measured | 67.1% | Only Qwen3.8 Max (0902) measured |
| SAGE (Vals AI) | 35.0% | 51.2% | Qwen3.8 Max (0902) |
| MedScribe (Vals AI) | 86.9% | 85.0% | Grok 4.5 |
| Vals Multimodal Index | 63.4% | 65.4% | Qwen3.8 Max (0902) |
| Long context reasoning (AA-LCR) | 79.3% | 80.3% | Qwen3.8 Max (0902) |
| DTBench | 96.5% | 92.0% | Grok 4.5 |
| LMCA | 45.2% | 46.2% | Qwen3.8 Max (0902) |
| Surface Evolver Bench | 74.4% | Not measured | Only Grok 4.5 measured |
| Furniture assembly | 22.5% | 20.0% | Grok 4.5 |
| LMArena Text | 1448 | 1483 | Qwen3.8 Max (0902) |
| LMArena Text: instruction following | 1445 | 1481 | Qwen3.8 Max (0902) |
| LMArena Text: creative writing | 1442 | 1477 | Qwen3.8 Max (0902) |
| LMArena Vision | 1287 | 1313 | Qwen3.8 Max (0902) |
| LMArena Text: English | 1453 | 1491 | Qwen3.8 Max (0902) |
| LMArena Text: non-English | 1440 | 1473 | Qwen3.8 Max (0902) |
| LMArena Text: Chinese | 1495 | 1539 | Qwen3.8 Max (0902) |
| LMArena Text: German | 1448 | 1478 | Qwen3.8 Max (0902) |
| LMArena Text: Japanese | 1430 | 1473 | Qwen3.8 Max (0902) |
| LMArena Text: Korean | 1406 | 1460 | Qwen3.8 Max (0902) |
| LMArena Text: French | 1457 | 1498 | Qwen3.8 Max (0902) |
| LMArena Text: Spanish | 1451 | 1489 | Qwen3.8 Max (0902) |
| LMArena Text: Russian | 1449 | 1482 | Qwen3.8 Max (0902) |
| LMArena Text: Polish | 1432 | 1471 | Qwen3.8 Max (0902) |
| LMArena Text: multi-turn | 1456 | 1489 | Qwen3.8 Max (0902) |
| LMArena Text: longer queries | 1462 | 1488 | Qwen3.8 Max (0902) |
| LMArena Text: writing and language | 1445 | 1477 | Qwen3.8 Max (0902) |
| LMArena Text: entertainment and media | 1434 | 1464 | Qwen3.8 Max (0902) |
| LMArena Vision: English | 1285 | 1311 | Qwen3.8 Max (0902) |
| LMArena Vision: Chinese | 1345 | 1358 | Qwen3.8 Max (0902) |
| LMArena Vision: reading text | 1293 | 1322 | Qwen3.8 Max (0902) |
| LMArena Vision: humor | 1276 | 1304 | Qwen3.8 Max (0902) |
| LMArena Vision: creative writing | 1293 | 1312 | Qwen3.8 Max (0902) |
| LMArena Document | 1452 | Not measured | Only Grok 4.5 measured |
| Neither model has a result on the other 12 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Grok 4.5 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Same cost
Repository review
50K input + 3K output tokens
Same cost
Agent loop
200K input + 10K output tokens
Same cost
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | SpaceXAI Grok 4.5 | Qwen3.8 Max (0902) |
|---|---|---|
| Maker | xAI | Qwen |
| Weights | No public weights listed | No public weights listed |
| Released | July 8, 2026 | September 3, 2026 |
| Context window | 500K | 1M |
| Input $/Mtok | $2.00 | $2.00 |
| Output $/Mtok | $6.00 | $6.00 |
| Coding index | 75.5 | 63.5 |
| Independent measurements behind it | 2 | 2 |
| Cost per solved task | $0.62 | $1.32 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Grok 4.5 better than Qwen3.8 Max (0902) for coding?
Pick Grok 4.5 for coding: it scores 75.5 on the Vibe Coding Index against 63.5 for Qwen3.8 Max (0902). Pick Qwen3.8 Max (0902) when the context window is larger, 1M against 500K tokens, or when what matters is LMArena WebDev or AA Coding Index, where it leads.
Which is cheaper, Grok 4.5 or Qwen3.8 Max (0902)?
Grok 4.5 costs $2.00 per million input tokens and $6.00 per million output tokens; Qwen3.8 Max (0902) costs $2.00 and $6.00. Per solved SWE-bench Verified task, Grok 4.5 costs $0.62 and Qwen3.8 Max (0902) $1.32.
Which has the bigger context window, Grok 4.5 or Qwen3.8 Max (0902)?
Qwen3.8 Max (0902), with 1M tokens against 500K.
How many benchmarks is the Grok 4.5 vs Qwen3.8 Max (0902) comparison based on?
105 benchmarks have a result for both models as of October 21, 2026. Grok 4.5's coding index rests on 2 independent benchmarks and Qwen3.8 Max (0902)'s on 2. Where only one model has a result, the page says so instead of naming a winner.