Model comparison
Inkling vs Qwen3.5 122B-A10B
Inkling is the stronger model overall, scoring 44.1 to 42.1 on the Noometry Index. Qwen3.5 122B-A10B costs 2.3× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. Inkling scores higher in 6 categories and Qwen3.5 122B-A10B in 2 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Inkling leads 55.1 to 38.8.
- The biggest single-benchmark swing is SciCode: 47% for Inkling and 35.6% for Qwen3.5 122B-A10B.
- Qwen3.5 122B-A10B is cheaper at $0.40 / $3.20 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- Qwen3.5 122B-A10B accepts more context: 262K tokens versus 66K.
Side by side
| Inkling | Qwen3.5 122B-A10B | |
|---|---|---|
| Provider | Thinking Machines Lab | Alibaba (Qwen) |
| Noometry Index | 44.1 | 42.1 |
| Released | 2026-07-15 | 2026-02-23 |
| Weights | Open | Open |
| Context window | 66K | 262K |
| Max output | 66K | 66K |
| Input $ / M tokens | $1.87 | $0.40 |
| Output $ / M tokens | $4.68 | $3.20 |
| Results tracked | 41 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.5 122B-A10B leads
Inkling: 34.5 (#234), Qwen3.5 122B-A10B: 39.1 (#162)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena WebDev | 1413 | 1360 |
| SciCode | 47% | 35.6% |
| LMArena Coding | 1464 | 1436 |
| FrontierCode | 14% | — |
| FrontierSWE | 4.1% | — |
| WeirdML | 32.3% | — |
| ALE-Bench | 946 | — |
Agentic & Tool Use Not comparable
Inkling: 29.6 (#85), Qwen3.5 122B-A10B: —
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| APEX-Agents | 33.8% | — |
| τ²-bench Banking | 25% | — |
Reasoning Inkling leads
Inkling: 40.4 (#56), Qwen3.5 122B-A10B: 27.2 (#123)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| CritPt | 5.4% | 0.9% |
| LMArena Hard Prompts | 1451 | 1421 |
| DTBench | 87.5% | 84.3% |
| LMCA | 37.6% | 32.2% |
| ARC-AGI-2 | 36.5% | — |
| SimpleBench | 50% | — |
| NYT Connections (extended) | — | 51.7% |
| ARC-AGI-1 | 79.5% | — |
| Chess Puzzles | 21% | — |
| Thematic Generalization | — | 51.2% |
| Mystery Game Puzzles | — | 17% |
| Epoch Capabilities Index | 148.54 | — |
Math Qwen3.5 122B-A10B leads
Inkling: 31.3 (#225), Qwen3.5 122B-A10B: 39.1 (#112)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Math | 1479 | 1432 |
| FrontierMath (Tiers 1-3) | 33.3% | — |
| FrontierMath Tier 4 | 4.9% | — |
| OTIS Mock AIME 2024-2025 | 88.9% | — |
| ProofBench | 0% | — |
Knowledge Inkling leads
Inkling: 55.1 (#49), Qwen3.5 122B-A10B: 38.8 (#142)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Expert | 1465 | 1432 |
| GPQA Diamond | 88.3% | — |
| SimpleQA Verified | 40.3% | — |
| Vectara Hallucination Rate | — | 11.2% |
Multimodal Not comparable
Inkling: —, Qwen3.5 122B-A10B: 39.6 (#57)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Vision | — | 1245 |
Multilingual Inkling leads
Inkling: 54.0 (#52), Qwen3.5 122B-A10B: 51.6 (#107)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Non-English | 1434 | 1400 |
| LMArena Chinese | 1490 | 1462 |
| LMArena French | 1458 | 1442 |
| LMArena German | 1446 | 1426 |
| LMArena Japanese | 1429 | 1367 |
| LMArena Korean | 1404 | 1352 |
| LMArena Russian | 1429 | 1400 |
| LMArena Spanish | 1448 | 1424 |
Instruction Following Inkling leads
Inkling: 75.1 (#71), Qwen3.5 122B-A10B: 73.8 (#115)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Instruction Following | 1426 | 1399 |
Long Context Too close to call
Inkling: 43.8 (#86), Qwen3.5 122B-A10B: 43.0 (#109)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Longer Query | 1434 | 1410 |
Writing & Preference Inkling leads
Inkling: 65.2 (#51), Qwen3.5 122B-A10B: 60.0 (#105)
| Benchmark | Inkling | Qwen3.5 122B-A10B |
|---|---|---|
| LMArena Text | 1441 | 1417 |
| LMArena Creative Writing | 1387 | 1368 |
| LMArena Multi-Turn | 1436 | 1416 |
| EQ-Bench Creative Writing | 1611 | — |
| EQ-Bench 4 | 1226 | — |
Frequently asked questions
Is Inkling better than Qwen3.5 122B-A10B?
Inkling is the stronger model overall, scoring 44.1 to 42.1 on the Noometry Index. Qwen3.5 122B-A10B costs 2.3× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Inkling or Qwen3.5 122B-A10B?
Qwen3.5 122B-A10B is cheaper. It lists at $0.40 per million input tokens and $3.20 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Inkling or Qwen3.5 122B-A10B better for coding?
Qwen3.5 122B-A10B scores higher on coding benchmarks: 39.1 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
Qwen3.5 122B-A10B does, with 262K tokens against 66K.
How many benchmarks do Inkling and Qwen3.5 122B-A10B share?
22 benchmarks have published results for both models. Inkling has 41 scored results on Noometry and Qwen3.5 122B-A10B has 27.