Model comparison
Inkling-Small vs Kimi K3
Kimi K3 is the stronger model overall, scoring 59.5 to 46.5 on the Noometry Index. Inkling-Small costs 9.4× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. Inkling-Small scores higher in 0 categories and Kimi K3 in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Kimi K3 leads 74.0 to 44.9.
- The biggest single-benchmark swing is ProofBench: 6% for Inkling-Small and 87% for Kimi K3.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $3 / $15 for Kimi K3.
- Kimi K3 accepts more context: 1.05M tokens versus 524K.
Side by side
| Inkling-Small | Kimi K3 | |
|---|---|---|
| Provider | Thinking Machines Lab | Moonshot AI |
| Noometry Index | 46.5 | 59.5 |
| Released | 2026-07-15 | 2026-07-16 |
| Weights | Open | Open |
| Context window | 524K | 1.05M |
| Max output | 1.05M | 1.05M |
| Input $ / M tokens | $0.45 | $3 |
| Output $ / M tokens | $1.20 | $15 |
| Results tracked | 33 | 53 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Kimi K3 leads
Inkling-Small: 43.3 (#80), Kimi K3: 60.6 (#8)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena WebDev | 1408 | 1659 |
| SciCode | 48.7% | 59.5% |
| LMArena Coding | 1451 | 1509 |
| DeepSWE | — | 68.5% |
| FrontierCode | — | 44.2% |
| FrontierSWE | — | 25.9% |
| WeirdML | — | 82.6% |
| ALE-Bench | — | 1,524 |
Agentic & Tool Use Not comparable
Inkling-Small: —, Kimi K3: 41.8 (#19)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| APEX-Agents | — | 50.6% |
| τ²-bench Banking | — | 37.1% |
| PostTrainBench | — | 32% |
| GBAEval | — | 48.3% |
| GDP.pdf | — | 19% |
| Vending-Bench 2 | — | 5,165 |
Reasoning Kimi K3 leads
Inkling-Small: 39.0 (#53), Kimi K3: 63.0 (#15)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| ARC-AGI-2 | 40.1% | 60.4% |
| ARC-AGI-1 | 84% | 94.5% |
| CritPt | 8.3% | 23.4% |
| Chess Puzzles | 18% | 39% |
| LMArena Hard Prompts | 1424 | 1495 |
| Mystery Game Puzzles | 6% | 26% |
| Epoch Capabilities Index | 150.13 | 157.68 |
| SimpleBench | — | 60.7% |
| NYT Connections (extended) | — | 93.6% |
| DTBench | — | 91.2% |
| LMCA | — | 52.7% |
| Surface Evolver Bench | — | 95% |
| ForecastBench | — | 61.1 |
Math Kimi K3 leads
Inkling-Small: 44.9 (#65), Kimi K3: 74.0 (#13)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.3% | 72.2% |
| FrontierMath Tier 4 | 17.1% | 39% |
| OTIS Mock AIME 2024-2025 | 90% | 97.2% |
| ProofBench | 6% | 87% |
| LMArena Math | 1453 | 1493 |
| MathArena Final-Answer Competitions | — | 87.8% |
Knowledge Kimi K3 leads
Inkling-Small: 48.3 (#71), Kimi K3: 63.2 (#19)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| GPQA Diamond | 88.5% | 93.1% |
| SimpleQA Verified | 19.1% | 50.6% |
| LMArena Expert | 1442 | 1520 |
Multimodal Too close to call
Inkling-Small: 38.4 (#59), Kimi K3: 38.4 (#58)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena Vision | 1235 | — |
| Blueprint-Bench 2 | — | 29.5% |
| Furniture Assembly | — | 34.2% |
Multilingual Kimi K3 leads
Inkling-Small: 51.7 (#102), Kimi K3: 56.5 (#20)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena Non-English | 1400 | 1466 |
| LMArena Chinese | 1458 | 1528 |
| LMArena French | 1438 | 1488 |
| LMArena German | 1412 | 1474 |
| LMArena Japanese | 1399 | 1492 |
| LMArena Korean | 1355 | 1465 |
| LMArena Russian | 1392 | 1481 |
| LMArena Spanish | 1426 | 1478 |
Instruction Following Kimi K3 leads
Inkling-Small: 73.9 (#112), Kimi K3: 77.8 (#13)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena Instruction Following | 1397 | 1481 |
Long Context Kimi K3 leads
Inkling-Small: 42.7 (#113), Kimi K3: 45.8 (#27)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena Longer Query | 1401 | 1494 |
Writing & Preference Kimi K3 leads
Inkling-Small: 59.8 (#102), Kimi K3: 76.7 (#4)
| Benchmark | Inkling-Small | Kimi K3 |
|---|---|---|
| LMArena Text | 1413 | 1476 |
| LMArena Creative Writing | 1330 | 1452 |
| EQ-Bench Creative Writing | 1491 | 2082 |
| LMArena Multi-Turn | 1420 | 1488 |
| EQ-Bench 4 | — | 1339 |
Frequently asked questions
Is Inkling-Small better than Kimi K3?
Kimi K3 is the stronger model overall, scoring 59.5 to 46.5 on the Noometry Index. Inkling-Small costs 9.4× less per token, which makes it the better buy when Kimi K3's lead doesn't matter for your workload.
Which is cheaper, Inkling-Small or Kimi K3?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; Kimi K3 lists at $3 and $15.
Is Inkling-Small or Kimi K3 better for coding?
Kimi K3 scores higher on coding benchmarks: 60.6 versus 43.3 in the Noometry coding category.
Which has the bigger context window?
Kimi K3 does, with 1.05M tokens against 524K.
How many benchmarks do Inkling-Small and Kimi K3 share?
32 benchmarks have published results for both models. Inkling-Small has 33 scored results on Noometry and Kimi K3 has 53.