Qwen3.8 Max (0902) vs Inkling
Updated October 21, 2026. 92 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Qwen3.8 Max (0902) for coding: it scores 63.5 on the Vibe Coding Index against 49.7 for Inkling.
Pick Inkling when output tokens cost 1.5x less, or when a solved task costs $0.76 against $1.32.
- Shared results
- 92
- Qwen3.8 Max (0902) only
- 21
- Inkling only
- 10
- 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
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) scores 63.5 against 49.7.
2 benchmarks behind each score
Intelligence
AA Intelligence Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 45.4 against 25.0.
Same board, both measured
Agentic work
AA Agentic Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 56.0 against 22.5.
Same board, both measured
Reasoning
GPQA Diamond
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 93.7% against 87.1%.
Same board, both measured
Knowledge
MMLU Pro
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 88.6% against 86.3%.
Same board, both measured
Web design
Design Arena: websites
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 1275 against 1224.
Same board, both measured
General
Vals Index
No call
Only Inkling has a Vals Index result, 29.0%.
Directional only
Long documents
Listed context window
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) takes 1M tokens against 524K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Inkling
Qwen3.8 Max (0902) $0.0050, Inkling $0.0030 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
Inkling
Qwen3.8 Max (0902) $1.32, Inkling $0.76.
Published attempt cost and pass rate
At a glance
- Coding index
- Qwen3.8 Max (0902) 63.5 from 2 benchmarks, Inkling 49.7 from 2 benchmarks.
- Widest category
- Math, 26.8 points apart on average over 6 shared benchmarks.
- Widest single gap
- ProofBench: Qwen3.8 Max (0902) 58.0% against Inkling 0.0%, 58.0 points apart on a 0 to 100 scale.
- Output price
- Qwen3.8 Max (0902) $6.00 against Inkling $4.05 per million output tokens, 1.5x apart.
- Context window
- Qwen3.8 Max (0902) takes 1M tokens, Inkling 524K.
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 | Qwen3.8 Max (0902) | Bars | Inkling | Ties |
|---|---|---|---|---|
| Coding | 17 | 2 | 1 tie | |
| Intelligence | 2 | 0 | ||
| Agentic work | 10 | 0 | ||
| Reasoning | 10 | 0 | ||
| Math | 6 | 0 | ||
| Knowledge | 8 | 2 | 1 tie | |
| Web design | 8 | 0 | ||
| General | 24 | 1 | ||
| All shared boards | 85 | 5 | 2 ties |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index172
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| LiveBench coding | 60.3% | 64.0% | Inkling |
| LMArena WebDev | 1674 | 1414 | Qwen3.8 Max (0902) |
| AA Coding Index | 76.2 | 52.1 | Qwen3.8 Max (0902) |
| LiveCodeBench (Vals AI) | 87.8% | 85.5% | Qwen3.8 Max (0902) |
| SWE-bench Verified (Vals AI) | 85.6% | 77.6% | Qwen3.8 Max (0902) |
| Terminal-Bench 4 (Vals AI) | 34.3% | 0.5% | Qwen3.8 Max (0902) |
| Vibe Code Bench (Vals AI) | 64.7% | 19.2% | Qwen3.8 Max (0902) |
| Code migration (Vals AI) | 24.0% | 11.8% | Qwen3.8 Max (0902) |
| ProgramBench (Vals AI) | 0.0% | 0.0% | Tie |
| IOI (Vals AI) | 68.9% | 14.9% | Qwen3.8 Max (0902) |
| Terminal-Bench 2.1 (Vals AI) | 67.4% | 47.6% | Qwen3.8 Max (0902) |
| SWE-bench Multimodal (Vals AI) | 42.5% | Not measured | Only Qwen3.8 Max (0902) measured |
| Vibe Code Bench 1-100 (Vals AI) | 12.8% | 7.3% | Qwen3.8 Max (0902) |
| SkillsBench (Vals AI) | 42.0% | 26.1% | Qwen3.8 Max (0902) |
| CyberBench (Vals AI) | 28.6% | 66.8% | Inkling |
| SciCode | 53.2% | 47.0% | Qwen3.8 Max (0902) |
| FrontierCode | Not measured | 14.0% | Only Inkling measured |
| FrontierSWE | 17.8% | 4.1% | Qwen3.8 Max (0902) |
| WeirdML | Not measured | 32.3% | Only Inkling measured |
| ALE-Bench | Not measured | 946.0 | Only Inkling measured |
| LMArena Text: coding | 1502 | 1466 | Qwen3.8 Max (0902) |
| LMArena Text: software and IT | 1498 | 1465 | Qwen3.8 Max (0902) |
| LMArena WebDev: React | 1692 | 1404 | Qwen3.8 Max (0902) |
| LMArena WebDev: HTML | 1649 | 1399 | Qwen3.8 Max (0902) |
| LMArena WebDev: from an image | 1637 | Not measured | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 13 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| AA Intelligence Index | 45.4 | 25.0 | Qwen3.8 Max (0902) |
| Epoch Capabilities Index | 157.4 | 148.4 | Qwen3.8 Max (0902) |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index100
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| AA Agentic Index | 56.0 | 22.5 | Qwen3.8 Max (0902) |
| Design Arena: agents building web apps | 1301 | 1196 | Qwen3.8 Max (0902) |
| Design Arena: agents building full-stack apps | 1305 | Not measured | Only Qwen3.8 Max (0902) measured |
| Design Arena: agents building mobile apps | 1223 | Not measured | Only Qwen3.8 Max (0902) measured |
| Design Arena: agents building Android apps | 1207 | Not measured | Only Qwen3.8 Max (0902) measured |
| Design Arena: agents building Godot games | 1234 | Not measured | Only Qwen3.8 Max (0902) measured |
| Design Arena: PowerPoint slides | 1257 | Not measured | Only Qwen3.8 Max (0902) measured |
| Design Arena: spreadsheets | 868 | Not measured | Only Qwen3.8 Max (0902) measured |
| LiveBench research | 69.9% | 59.9% | Qwen3.8 Max (0902) |
| LiveBench browser | 32.0% | 16.0% | Qwen3.8 Max (0902) |
| LiveBench long agentic | 42.3% | 14.1% | Qwen3.8 Max (0902) |
| SREBench (Vals AI) | Not measured | 0.0% | Only Inkling measured |
| Finance Agent (Vals AI) | 50.6% | 46.6% | Qwen3.8 Max (0902) |
| Tax Agent Bench (Vals AI) | 32.1% | 15.3% | Qwen3.8 Max (0902) |
| Legal Research Bench (Vals AI) | 47.6% | 28.4% | Qwen3.8 Max (0902) |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 3.8% | Only Inkling measured |
| Excel Modeling Benchmark (Vals AI) | 60.1% | 38.4% | Qwen3.8 Max (0902) |
| Terminal-Bench-Science (Vals AI) | 12.9% | Not measured | Only Qwen3.8 Max (0902) measured |
| APEX-Agents | 63.3% | 33.8% | Qwen3.8 Max (0902) |
| LMArena Search | 1261 | Not measured | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 19 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index100
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| GPQA Diamond | 93.7% | 87.1% | Qwen3.8 Max (0902) |
| LiveBench reasoning | 88.2% | 78.3% | Qwen3.8 Max (0902) |
| MysteryMechanism (Vals AI) | 23.9% | Not measured | Only Qwen3.8 Max (0902) measured |
| MMMU Pro (Vals AI) | 88.0% | 78.5% | Qwen3.8 Max (0902) |
| GPQA Diamond (Artificial Analysis run) | 92.8% | 87.2% | Qwen3.8 Max (0902) |
| Humanity's Last Exam (Artificial Analysis run) | 43.1% | 31.9% | Qwen3.8 Max (0902) |
| GPQA Diamond (Epoch AI run) | 92.7% | 88.3% | Qwen3.8 Max (0902) |
| ARC-AGI | Not measured | 79.5% | Only Inkling measured |
| ARC-AGI-2 | Not measured | 36.5% | Only Inkling measured |
| SimpleBench | Not measured | 50.0% | Only Inkling measured |
| Chess puzzles | 46.0% | 21.0% | Qwen3.8 Max (0902) |
| Mystery game puzzles | 38.0% | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Text: hard prompts | 1496 | 1451 | Qwen3.8 Max (0902) |
| LMArena Text: expert | 1507 | 1466 | Qwen3.8 Max (0902) |
| LMArena Text: hard prompts in English | 1503 | 1448 | Qwen3.8 Max (0902) |
| LMArena Vision: diagrams | 1323 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Vision: homework | 1336 | Not measured | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 88.9% | Qwen3.8 Max (0902) |
| FrontierMath, tiers 1 to 3 | 74.7% | 33.3% | Qwen3.8 Max (0902) |
| FrontierMath, tier 4 | 46.3% | 4.9% | Qwen3.8 Max (0902) |
| ProofBench | 58.0% | 0.0% | Qwen3.8 Max (0902) |
| LMArena Text: math | 1501 | 1479 | Qwen3.8 Max (0902) |
| LMArena Text: mathematical work | 1491 | 1460 | 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 index82
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| MMLU Pro | 88.6% | 86.3% | Qwen3.8 Max (0902) |
| LegalBench (Vals AI) | 83.6% | 83.0% | Qwen3.8 Max (0902) |
| CorpFin (Vals AI) | 65.8% | 68.6% | Inkling |
| TaxEval (Vals AI) | 75.5% | 75.3% | Qwen3.8 Max (0902) |
| MortgageTax (Vals AI) | 64.0% | 64.0% | Tie |
| MedCode (Vals AI) | 40.7% | 41.2% | Inkling |
| SimpleQA Verified | 47.3% | 40.3% | Qwen3.8 Max (0902) |
| LMArena Text: science | 1512 | 1455 | Qwen3.8 Max (0902) |
| LMArena Text: medicine and healthcare | 1503 | 1447 | Qwen3.8 Max (0902) |
| LMArena Text: legal and government | 1492 | 1448 | Qwen3.8 Max (0902) |
| LMArena Text: business and finance | 1480 | 1438 | 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 index80
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| Design Arena: websites | 1275 | 1224 | Qwen3.8 Max (0902) |
| Design Arena: UI components | 1340 | 1200 | Qwen3.8 Max (0902) |
| Design Arena: game dev | 1313 | 1166 | Qwen3.8 Max (0902) |
| Design Arena: data viz | 1281 | 1160 | Qwen3.8 Max (0902) |
| Design Arena: 3D | 1337 | 1161 | Qwen3.8 Max (0902) |
| Design Arena: all code categories | 1293 | 1206 | Qwen3.8 Max (0902) |
| Design Arena: SVG | Not measured | 1109 | Only Inkling measured |
| Design Arena: ASCII art | 1201 | 1089 | Qwen3.8 Max (0902) |
| LiveBench design | 78.6% | 58.3% | Qwen3.8 Max (0902) |
GeneralLike-for-likePublic boards, shown as published, kept out of the index241
| Benchmark | Qwen3.8 Max (0902) | Inkling | Leader |
|---|---|---|---|
| Vals Index | Not measured | 29.0% | Only Inkling measured |
| LiveBench data analysis | 78.4% | 72.8% | Qwen3.8 Max (0902) |
| Public Benefits Bench (Vals AI) | 67.1% | 58.5% | Qwen3.8 Max (0902) |
| SAGE (Vals AI) | 51.2% | 36.5% | Qwen3.8 Max (0902) |
| MedScribe (Vals AI) | 85.0% | 85.4% | Inkling |
| Vals Multimodal Index | 65.4% | 49.4% | Qwen3.8 Max (0902) |
| Long context reasoning (AA-LCR) | 80.3% | 77.3% | Qwen3.8 Max (0902) |
| DTBench | 92.0% | 87.5% | Qwen3.8 Max (0902) |
| LMCA | 46.2% | 37.6% | Qwen3.8 Max (0902) |
| Furniture assembly | 20.0% | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Text | 1483 | 1442 | Qwen3.8 Max (0902) |
| LMArena Text: instruction following | 1481 | 1425 | Qwen3.8 Max (0902) |
| LMArena Text: creative writing | 1477 | 1390 | Qwen3.8 Max (0902) |
| LMArena Vision | 1313 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Text: English | 1491 | 1442 | Qwen3.8 Max (0902) |
| LMArena Text: non-English | 1473 | 1434 | Qwen3.8 Max (0902) |
| LMArena Text: Chinese | 1539 | 1493 | Qwen3.8 Max (0902) |
| LMArena Text: German | 1478 | 1445 | Qwen3.8 Max (0902) |
| LMArena Text: Japanese | 1473 | 1429 | Qwen3.8 Max (0902) |
| LMArena Text: Korean | 1460 | 1403 | Qwen3.8 Max (0902) |
| LMArena Text: French | 1498 | 1455 | Qwen3.8 Max (0902) |
| LMArena Text: Spanish | 1489 | 1449 | Qwen3.8 Max (0902) |
| LMArena Text: Russian | 1482 | 1430 | Qwen3.8 Max (0902) |
| LMArena Text: Polish | 1471 | 1450 | Qwen3.8 Max (0902) |
| LMArena Text: multi-turn | 1489 | 1437 | Qwen3.8 Max (0902) |
| LMArena Text: longer queries | 1488 | 1435 | Qwen3.8 Max (0902) |
| LMArena Text: writing and language | 1477 | 1404 | Qwen3.8 Max (0902) |
| LMArena Text: entertainment and media | 1464 | 1394 | Qwen3.8 Max (0902) |
| LMArena Vision: English | 1311 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Vision: Chinese | 1358 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Vision: reading text | 1322 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Vision: humor | 1304 | Not measured | Only Qwen3.8 Max (0902) measured |
| LMArena Vision: creative writing | 1312 | Not measured | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 13 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen3.8 Max (0902) | Inkling | 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
Inkling costs less
Repository review
50K input + 3K output tokens
Inkling costs less
Agent loop
200K input + 10K output tokens
Inkling costs less
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 | Qwen3.8 Max (0902) | Thinking Machines Inkling |
|---|---|---|
| Maker | Qwen | Thinking Machines |
| Weights | No public weights listed | Open weights |
| Released | September 3, 2026 | July 17, 2026 |
| Context window | 1M | 524K |
| Input $/Mtok | $2.00 | $1.00 |
| Output $/Mtok | $6.00 | $4.05 |
| Coding index | 63.5 | 49.7 |
| Independent measurements behind it | 2 | 2 |
| Cost per solved task | $1.32 | $0.76 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen3.8 Max (0902) better than Inkling for coding?
Pick Qwen3.8 Max (0902) for coding: it scores 63.5 on the Vibe Coding Index against 49.7 for Inkling. Pick Inkling when output tokens cost 1.5x less, or when a solved task costs $0.76 against $1.32.
Which is cheaper, Qwen3.8 Max (0902) or Inkling?
Qwen3.8 Max (0902) costs $2.00 per million input tokens and $6.00 per million output tokens; Inkling costs $1.00 and $4.05. Per solved SWE-bench Verified task, Qwen3.8 Max (0902) costs $1.32 and Inkling $0.76.
Which has the bigger context window, Qwen3.8 Max (0902) or Inkling?
Qwen3.8 Max (0902), with 1M tokens against 524K.
How many benchmarks is the Qwen3.8 Max (0902) vs Inkling comparison based on?
92 benchmarks have a result for both models as of October 21, 2026. Qwen3.8 Max (0902)'s coding index rests on 2 independent benchmarks and Inkling's on 2. Where only one model has a result, the page says so instead of naming a winner.