Model comparison
GPT-4.1 vs Inkling-Small
Inkling-Small is the stronger model overall, scoring 46.5 to 35.9 on the Noometry Index.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. GPT-4.1 scores higher in 0 categories and Inkling-Small in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling-Small leads 38.6 to 11.7.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for GPT-4.1 and 84% for Inkling-Small.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $2 / $8 for GPT-4.1.
- GPT-4.1 accepts more context: 1.05M tokens versus 524K.
- Inkling-Small has downloadable open weights; the other is API-only.
Side by side
| GPT-4.1 | Inkling-Small | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 35.9 | 46.5 |
| Released | 2025-04-14 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 1.05M | 524K |
| Max output | 33K | 1.05M |
| Input $ / M tokens | $2 | $0.45 |
| Output $ / M tokens | $8 | $1.20 |
| Results tracked | 52 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
GPT-4.1: 34.4 (#238), Inkling-Small: 43.6 (#85)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Coding | 1391 | 1451 |
| SWE-bench Verified | 48.5% | — |
| SWE-bench Verified (bash only) | 39.6% | — |
| Aider Polyglot | 52.4% | — |
| LMArena WebDev | — | 1409 |
| SciCode | — | 48.7% |
| WeirdML | 39% | — |
| CadEval | 42% | — |
| ALE-Bench | 558.1 | — |
Agentic & Tool Use Not comparable
GPT-4.1: 34.7 (#43), Inkling-Small: —
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| Berkeley Function Calling Leaderboard | 54% | — |
Reasoning Inkling-Small leads
GPT-4.1: 11.7 (#339), Inkling-Small: 38.6 (#63)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| ARC-AGI-2 | 0.4% | 40.1% |
| ARC-AGI-1 | 5.5% | 84% |
| Chess Puzzles | 6% | 18% |
| LMArena Hard Prompts | 1384 | 1423 |
| Epoch Capabilities Index | 136.78 | 150.15 |
| SimpleBench | 27% | — |
| Kagi LLM Benchmark | 52.3% | — |
| CritPt | — | 8.3% |
| EnigmaEval | 2.2% | — |
| Mystery Game Puzzles | — | 6% |
| DTBench | 68.3% | — |
| LMCA | 25.6% | — |
| ForecastBench | 61.5 | — |
Math Inkling-Small leads
GPT-4.1: 22.3 (#280), Inkling-Small: 45.1 (#77)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | 6% | 46.3% |
| OTIS Mock AIME 2024-2025 | 38.3% | 90% |
| LMArena Math | 1370 | 1459 |
| FrontierMath Tier 4 | — | 17.1% |
| ProofBench | — | 6% |
| Omni-MATH | 47.1% | — |
| MATH Level 5 | 83% | — |
| FrontierMath (Feb 2025 set) | 5.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Inkling-Small leads
GPT-4.1: 37.1 (#160), Inkling-Small: 48.2 (#77)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| GPQA Diamond | 66.9% | 88.5% |
| SimpleQA Verified | 31.1% | 19.1% |
| LMArena Expert | 1364 | 1442 |
| Humanity's Last Exam | 5.4% | — |
| MMLU-Pro | 81.1% | — |
| Vectara Hallucination Rate | 5.6% | — |
| GPQA (HELM) | 65.9% | — |
Multimodal Too close to call
GPT-4.1: 38.2 (#67), Inkling-Small: 39.1 (#62)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Vision | 1211 | 1235 |
| GeoBench | 72% | — |
Multilingual Inkling-Small leads
GPT-4.1: 49.4 (#133), Inkling-Small: 51.7 (#104)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1370 | 1402 |
| LMArena Chinese | 1382 | 1465 |
| LMArena French | 1382 | 1436 |
| LMArena German | 1381 | 1405 |
| LMArena Japanese | 1319 | 1405 |
| LMArena Korean | 1339 | 1363 |
| LMArena Russian | 1377 | 1391 |
| LMArena Spanish | 1376 | 1428 |
Instruction Following Inkling-Small leads
GPT-4.1: 71.3 (#153), Inkling-Small: 73.8 (#114)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1367 | 1399 |
| IFEval | 83.8% | — |
Long Context Inkling-Small leads
GPT-4.1: 40.0 (#163), Inkling-Small: 42.7 (#118)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1385 | 1401 |
| Fiction.LiveBench | 63.9% | — |
Writing & Preference Inkling-Small leads
GPT-4.1: 57.6 (#125), Inkling-Small: 59.6 (#107)
| Benchmark | GPT-4.1 | Inkling-Small |
|---|---|---|
| LMArena Text | 1383 | 1414 |
| LMArena Creative Writing | 1363 | 1331 |
| EQ-Bench Creative Writing | 1420 | 1491 |
| LMArena Multi-Turn | 1398 | 1418 |
| WildBench | 85.4% | — |
Frequently asked questions
Is GPT-4.1 better than Inkling-Small?
Inkling-Small is the stronger model overall, scoring 46.5 to 35.9 on the Noometry Index.
Which is cheaper, GPT-4.1 or Inkling-Small?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; GPT-4.1 lists at $2 and $8.
Is GPT-4.1 or Inkling-Small better for coding?
Inkling-Small scores higher on coding benchmarks: 43.6 versus 34.4 in the Noometry coding category.
Which has the bigger context window?
GPT-4.1 does, with 1.05M tokens against 524K.
How many benchmarks do GPT-4.1 and Inkling-Small share?
27 benchmarks have published results for both models. GPT-4.1 has 52 scored results on Noometry and Inkling-Small has 33.