Model comparison
Inkling vs Qwen2.5 7B Instruct
Inkling is the stronger model overall, scoring 44.1 to 29.0 on the Noometry Index. Qwen2.5 7B Instruct costs 8.4× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 6 shared benchmarks.
Alibaba (Qwen)Summary
- They share 6 benchmarks with published results for both. Inkling scores higher in 6 categories and Qwen2.5 7B Instruct in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Inkling leads 55.1 to 17.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 88.9% for Inkling and 2.5% for Qwen2.5 7B Instruct.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- Qwen2.5 7B Instruct accepts more context: 131K tokens versus 66K.
Side by side
| Inkling | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | Thinking Machines Lab | Alibaba (Qwen) |
| Noometry Index | 44.1 | 29.0 |
| Released | 2026-07-15 | 2024-09 |
| Weights | Open | Open |
| Context window | 66K | 131K |
| Max output | 66K | 8K |
| Input $ / M tokens | $1.87 | $0.17 |
| Output $ / M tokens | $4.68 | $0.70 |
| Results tracked | 41 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen2.5 7B Instruct leads
Inkling: 34.5 (#234), Qwen2.5 7B Instruct: 36.5 (#208)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| FrontierCode | 14% | — |
| LMArena WebDev | 1413 | — |
| FrontierSWE | 4.1% | — |
| SciCode | 47% | — |
| WeirdML | 32.3% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1464 | — |
| BigCodeBench Complete | — | 46.1% |
| ALE-Bench | 946 | — |
Agentic & Tool Use Inkling leads
Inkling: 29.6 (#85), Qwen2.5 7B Instruct: 23.8 (#124)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| APEX-Agents | 33.8% | — |
| τ²-bench Banking | 25% | — |
| BALROG | — | 7.8% |
Reasoning Inkling leads
Inkling: 40.4 (#56), Qwen2.5 7B Instruct: 14.8 (#322)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| Chess Puzzles | 21% | 0% |
| DTBench | 87.5% | 47.7% |
| LMCA | 37.6% | 6.4% |
| Epoch Capabilities Index | 148.54 | 118.51 |
| ARC-AGI-2 | 36.5% | — |
| SimpleBench | 50% | — |
| ARC-AGI-1 | 79.5% | — |
| CritPt | 5.4% | — |
| LMArena Hard Prompts | 1451 | — |
Math Inkling leads
Inkling: 31.3 (#225), Qwen2.5 7B Instruct: 12.6 (#306)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.9% | 2.5% |
| FrontierMath (Tiers 1-3) | 33.3% | — |
| FrontierMath Tier 4 | 4.9% | — |
| ProofBench | 0% | — |
| Omni-MATH | — | 29.4% |
| LMArena Math | 1479 | — |
Knowledge Inkling leads
Inkling: 55.1 (#49), Qwen2.5 7B Instruct: 17.0 (#286)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 88.3% | 35.5% |
| SimpleQA Verified | 40.3% | — |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1465 | — |
| MMLU | — | 72.9% |
Multilingual Not comparable
Inkling: 54.0 (#52), Qwen2.5 7B Instruct: —
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1434 | — |
| LMArena Chinese | 1490 | — |
| LMArena French | 1458 | — |
| LMArena German | 1446 | — |
| LMArena Japanese | 1429 | — |
| LMArena Korean | 1404 | — |
| LMArena Russian | 1429 | — |
| LMArena Spanish | 1448 | — |
Instruction Following Inkling leads
Inkling: 75.1 (#71), Qwen2.5 7B Instruct: 63.2 (#231)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1426 | — |
Long Context Not comparable
Inkling: 43.8 (#86), Qwen2.5 7B Instruct: —
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1434 | — |
Writing & Preference Inkling leads
Inkling: 65.2 (#51), Qwen2.5 7B Instruct: 48.8 (#195)
| Benchmark | Inkling | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1441 | — |
| LMArena Creative Writing | 1387 | — |
| EQ-Bench Creative Writing | 1611 | — |
| WildBench | — | 73.1% |
| EQ-Bench 4 | 1226 | — |
| LMArena Multi-Turn | 1436 | — |
Frequently asked questions
Is Inkling better than Qwen2.5 7B Instruct?
Inkling is the stronger model overall, scoring 44.1 to 29.0 on the Noometry Index. Qwen2.5 7B Instruct costs 8.4× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Inkling or Qwen2.5 7B Instruct?
Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Inkling or Qwen2.5 7B Instruct better for coding?
Qwen2.5 7B Instruct scores higher on coding benchmarks: 36.5 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
Qwen2.5 7B Instruct does, with 131K tokens against 66K.
How many benchmarks do Inkling and Qwen2.5 7B Instruct share?
6 benchmarks have published results for both models. Inkling has 41 scored results on Noometry and Qwen2.5 7B Instruct has 15.