Model comparison
GPT-5 Mini vs Inkling-Small
Inkling-Small is the stronger model overall, scoring 46.5 to 41.7 on the Noometry Index.
Last verified . 32 shared benchmarks.
Summary
- They share 32 benchmarks with published results for both. GPT-5 Mini scores higher in 2 categories and Inkling-Small in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling-Small leads 39.0 to 24.1.
- The biggest single-benchmark swing is ARC-AGI-2: 4.4% for GPT-5 Mini and 40.1% for Inkling-Small.
- Inkling-Small is cheaper at $0.45 / $1.20 per million input/output tokens, against $0.25 / $2 for GPT-5 Mini.
- Inkling-Small accepts more context: 524K tokens versus 400K.
- Inkling-Small has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | Inkling-Small | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 41.7 | 46.5 |
| Released | 2025-08-07 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 400K | 524K |
| Max output | 128K | 1.05M |
| Input $ / M tokens | $0.25 | $0.45 |
| Output $ / M tokens | $2 | $1.20 |
| Results tracked | 60 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling-Small leads
GPT-5 Mini: 40.0 (#134), Inkling-Small: 43.3 (#80)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| SciCode | 39.2% | 48.7% |
| LMArena Coding | 1406 | 1451 |
| SWE-bench Verified | 64.7% | — |
| SWE-bench Verified (bash only) | 59.8% | — |
| LMArena WebDev | — | 1408 |
| SWE-bench Multilingual | 39.7% | — |
| WeirdML | 52.7% | — |
| ALE-Bench | 799.77 | — |
| AlgoTune | 1.38 | — |
Agentic & Tool Use Not comparable
GPT-5 Mini: 31.0 (#64), Inkling-Small: —
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| Terminal-Bench | 34.8% | — |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| Vending-Bench 2 | -31.18 | — |
Reasoning Inkling-Small leads
GPT-5 Mini: 24.1 (#153), Inkling-Small: 39.0 (#53)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| ARC-AGI-2 | 4.4% | 40.1% |
| ARC-AGI-1 | 54.3% | 84% |
| CritPt | 0% | 8.3% |
| Chess Puzzles | 30% | 18% |
| LMArena Hard Prompts | 1380 | 1424 |
| Mystery Game Puzzles | 10% | 6% |
| Epoch Capabilities Index | 145.51 | 150.13 |
| Kagi LLM Benchmark | 70.3% | — |
| EnigmaEval | 8.2% | — |
| DTBench | 80.5% | — |
| LMCA | 34.2% | — |
| ForecastBench | 61 | — |
Math GPT-5 Mini leads
GPT-5 Mini: 46.3 (#59), Inkling-Small: 44.9 (#65)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.7% | 46.3% |
| FrontierMath Tier 4 | 12.2% | 17.1% |
| OTIS Mock AIME 2024-2025 | 86.7% | 90% |
| ProofBench | 9% | 6% |
| LMArena Math | 1377 | 1453 |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Inkling-Small leads
GPT-5 Mini: 45.5 (#80), Inkling-Small: 48.3 (#71)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| GPQA Diamond | 75% | 88.5% |
| SimpleQA Verified | 21.6% | 19.1% |
| LMArena Expert | 1377 | 1442 |
| Humanity's Last Exam | 19.4% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Inkling-Small leads
GPT-5 Mini: 35.0 (#80), Inkling-Small: 38.4 (#59)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| LMArena Vision | 1202 | 1235 |
| VPCT | 40.2% | — |
Multilingual Inkling-Small leads
GPT-5 Mini: 49.1 (#131), Inkling-Small: 51.7 (#102)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| LMArena Non-English | 1363 | 1400 |
| LMArena Chinese | 1387 | 1458 |
| LMArena French | 1385 | 1438 |
| LMArena German | 1366 | 1412 |
| LMArena Japanese | 1342 | 1399 |
| LMArena Korean | 1306 | 1355 |
| LMArena Russian | 1362 | 1392 |
| LMArena Spanish | 1355 | 1426 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.4 (#43), Inkling-Small: 73.9 (#112)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| LMArena Instruction Following | 1357 | 1397 |
| IFEval | 92.7% | — |
Long Context Too close to call
GPT-5 Mini: 41.9 (#127), Inkling-Small: 42.7 (#113)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| LMArena Longer Query | 1355 | 1401 |
| Fiction.LiveBench | 69.4% | — |
Writing & Preference Inkling-Small leads
GPT-5 Mini: 55.4 (#143), Inkling-Small: 59.8 (#102)
| Benchmark | GPT-5 Mini | Inkling-Small |
|---|---|---|
| LMArena Text | 1373 | 1413 |
| LMArena Creative Writing | 1325 | 1330 |
| EQ-Bench Creative Writing | 1313 | 1491 |
| LMArena Multi-Turn | 1363 | 1420 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |
Frequently asked questions
Is GPT-5 Mini better than Inkling-Small?
Inkling-Small is the stronger model overall, scoring 46.5 to 41.7 on the Noometry Index.
Which is cheaper, GPT-5 Mini or Inkling-Small?
Inkling-Small is cheaper. It lists at $0.45 per million input tokens and $1.20 per million output tokens; GPT-5 Mini lists at $0.25 and $2.
Is GPT-5 Mini or Inkling-Small better for coding?
Inkling-Small scores higher on coding benchmarks: 43.3 versus 40.0 in the Noometry coding category.
Which has the bigger context window?
Inkling-Small does, with 524K tokens against 400K.
How many benchmarks do GPT-5 Mini and Inkling-Small share?
32 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Inkling-Small has 33.