Model comparison
Inkling vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 44.1 on the Noometry Index.
Last verified . 33 shared benchmarks.
Alibaba (Qwen)Summary
- They share 33 benchmarks with published results for both. Inkling scores higher in 0 categories and Qwen3.8 Max in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.2 to 31.3.
- The biggest single-benchmark swing is ProofBench: 0% for Inkling and 58% for Qwen3.8 Max.
- Inkling is cheaper at $1.87 / $4.68 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Qwen3.8 Max accepts more context: 1M tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| Inkling | Qwen3.8 Max | |
|---|---|---|
| Provider | Thinking Machines Lab | Alibaba (Qwen) |
| Noometry Index | 44.1 | 56.8 |
| Released | 2026-07-15 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 66K | 1M |
| Max output | 66K | 131K |
| Input $ / M tokens | $1.87 | $2 |
| Output $ / M tokens | $4.68 | $6 |
| Results tracked | 41 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Inkling: 34.5 (#234), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena WebDev | 1413 | 1674 |
| FrontierSWE | 4.1% | 17.8% |
| SciCode | 47% | 53.2% |
| LMArena Coding | 1464 | 1502 |
| DeepSWE | — | 57.5% |
| FrontierCode | 14% | — |
| WeirdML | 32.3% | — |
| ALE-Bench | 946 | — |
Agentic & Tool Use Qwen3.8 Max leads
Inkling: 29.6 (#85), Qwen3.8 Max: 45.4 (#14)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | 33.8% | 63.3% |
| τ²-bench Banking | 25% | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
Inkling: 40.4 (#56), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| CritPt | 5.4% | 20% |
| Chess Puzzles | 21% | 40% |
| LMArena Hard Prompts | 1451 | 1496 |
| DTBench | 87.5% | 92% |
| LMCA | 37.6% | 46.2% |
| Epoch Capabilities Index | 148.54 | 156.41 |
| ARC-AGI-2 | 36.5% | — |
| SimpleBench | 50% | — |
| NYT Connections (extended) | — | 88.3% |
| ARC-AGI-1 | 79.5% | — |
| Mystery Game Puzzles | — | 38% |
Math Qwen3.8 Max leads
Inkling: 31.3 (#225), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| FrontierMath (Tiers 1-3) | 33.3% | 74.7% |
| FrontierMath Tier 4 | 4.9% | 46.3% |
| OTIS Mock AIME 2024-2025 | 88.9% | 100% |
| ProofBench | 0% | 58% |
| LMArena Math | 1479 | 1499 |
Knowledge Qwen3.8 Max leads
Inkling: 55.1 (#49), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 88.3% | 92.7% |
| SimpleQA Verified | 40.3% | 47.3% |
| LMArena Expert | 1465 | 1507 |
Multimodal Not comparable
Inkling: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
Inkling: 54.0 (#52), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1434 | 1472 |
| LMArena Chinese | 1490 | 1538 |
| LMArena French | 1458 | 1503 |
| LMArena German | 1446 | 1483 |
| LMArena Japanese | 1429 | 1467 |
| LMArena Korean | 1404 | 1461 |
| LMArena Russian | 1429 | 1481 |
| LMArena Spanish | 1448 | 1492 |
Instruction Following Qwen3.8 Max leads
Inkling: 75.1 (#71), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1426 | 1479 |
Long Context Qwen3.8 Max leads
Inkling: 43.8 (#86), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1434 | 1489 |
Writing & Preference Qwen3.8 Max leads
Inkling: 65.2 (#51), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Inkling | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1441 | 1483 |
| LMArena Creative Writing | 1387 | 1479 |
| LMArena Multi-Turn | 1436 | 1489 |
| EQ-Bench Creative Writing | 1611 | — |
| EQ-Bench 4 | 1226 | — |
Frequently asked questions
Is Inkling better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 44.1 on the Noometry Index.
Which is cheaper, Inkling or Qwen3.8 Max?
Inkling is cheaper. It lists at $1.87 per million input tokens and $4.68 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is Inkling or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 66K.
How many benchmarks do Inkling and Qwen3.8 Max share?
33 benchmarks have published results for both models. Inkling has 41 scored results on Noometry and Qwen3.8 Max has 39.