Model comparison
GPT-5 Mini vs Inkling
Inkling is the stronger model overall, scoring 44.1 to 41.8 on the Noometry Index. GPT-5 Mini costs 3.7× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 34 shared benchmarks.
Summary
- They share 34 benchmarks with published results for both. GPT-5 Mini scores higher in 4 categories and Inkling in 5 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling leads 40.4 to 23.9.
- The biggest single-benchmark swing is ARC-AGI-2: 4.4% for GPT-5 Mini and 36.5% for Inkling.
- GPT-5 Mini is cheaper at $0.25 / $2 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- GPT-5 Mini accepts more context: 400K tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| GPT-5 Mini | Inkling | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 41.8 | 44.1 |
| Released | 2025-08-07 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 400K | 66K |
| Max output | 128K | 66K |
| Input $ / M tokens | $0.25 | $1.87 |
| Output $ / M tokens | $2 | $4.68 |
| Results tracked | 60 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5 Mini leads
GPT-5 Mini: 40.1 (#146), Inkling: 34.5 (#234)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| SciCode | 39.2% | 47% |
| WeirdML | 52.7% | 32.3% |
| LMArena Coding | 1406 | 1464 |
| ALE-Bench | 799.77 | 946 |
| SWE-bench Verified | 64.7% | — |
| FrontierCode | — | 14% |
| SWE-bench Verified (bash only) | 59.8% | — |
| LMArena WebDev | — | 1413 |
| SWE-bench Multilingual | 39.7% | — |
| FrontierSWE | — | 4.1% |
| AlgoTune | 1.38 | — |
Agentic & Tool Use GPT-5 Mini leads
GPT-5 Mini: 31.1 (#70), Inkling: 29.6 (#85)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| Terminal-Bench | 34.8% | — |
| APEX-Agents | — | 33.8% |
| Berkeley Function Calling Leaderboard | 55.5% | — |
| τ²-bench Banking | — | 25% |
| Vending-Bench 2 | -31.18 | — |
Reasoning Inkling leads
GPT-5 Mini: 23.9 (#168), Inkling: 40.4 (#56)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| ARC-AGI-2 | 4.4% | 36.5% |
| ARC-AGI-1 | 54.3% | 79.5% |
| CritPt | 0% | 5.4% |
| Chess Puzzles | 30% | 21% |
| LMArena Hard Prompts | 1380 | 1451 |
| DTBench | 80.5% | 87.5% |
| LMCA | 34.2% | 37.6% |
| Epoch Capabilities Index | 145.52 | 148.54 |
| SimpleBench | — | 50% |
| Kagi LLM Benchmark | 70.3% | — |
| EnigmaEval | 8.2% | — |
| Mystery Game Puzzles | 10% | — |
| ForecastBench | 61 | — |
Math GPT-5 Mini leads
GPT-5 Mini: 46.7 (#69), Inkling: 31.3 (#225)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 46.7% | 33.3% |
| FrontierMath Tier 4 | 12.2% | 4.9% |
| OTIS Mock AIME 2024-2025 | 86.7% | 88.9% |
| ProofBench | 9% | 0% |
| LMArena Math | 1378 | 1479 |
| Omni-MATH | 72.2% | — |
| MATH Level 5 | 97.8% | — |
| FrontierMath (Feb 2025 set) | 27.2% | — |
| FrontierMath Tier 4 (v1) | 6.3% | — |
Knowledge Inkling leads
GPT-5 Mini: 45.6 (#86), Inkling: 55.1 (#49)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| GPQA Diamond | 75% | 88.3% |
| SimpleQA Verified | 21.6% | 40.3% |
| LMArena Expert | 1379 | 1465 |
| Humanity's Last Exam | 19.4% | — |
| MMLU-Pro | 83.5% | — |
| Confabulations | 13.3% | — |
| Vectara Hallucination Rate | 12.9% | — |
| GPQA (HELM) | 75.6% | — |
Multimodal Not comparable
GPT-5 Mini: 35.6 (#85), Inkling: —
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| LMArena Vision | 1202 | — |
| VPCT | 40.2% | — |
Multilingual Inkling leads
GPT-5 Mini: 48.9 (#137), Inkling: 54.0 (#52)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| LMArena Non-English | 1363 | 1434 |
| LMArena Chinese | 1385 | 1490 |
| LMArena French | 1386 | 1458 |
| LMArena German | 1366 | 1446 |
| LMArena Japanese | 1341 | 1429 |
| LMArena Korean | 1308 | 1404 |
| LMArena Russian | 1362 | 1429 |
| LMArena Spanish | 1355 | 1448 |
Instruction Following GPT-5 Mini leads
GPT-5 Mini: 76.2 (#46), Inkling: 75.1 (#71)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| LMArena Instruction Following | 1357 | 1426 |
| IFEval | 92.7% | — |
Long Context Inkling leads
GPT-5 Mini: 41.9 (#132), Inkling: 43.8 (#86)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| LMArena Longer Query | 1355 | 1434 |
| Fiction.LiveBench | 69.4% | — |
Writing & Preference Inkling leads
GPT-5 Mini: 55.2 (#148), Inkling: 65.2 (#51)
| Benchmark | GPT-5 Mini | Inkling |
|---|---|---|
| LMArena Text | 1373 | 1441 |
| LMArena Creative Writing | 1325 | 1387 |
| EQ-Bench Creative Writing | 1313 | 1611 |
| LMArena Multi-Turn | 1363 | 1436 |
| Short-Story Creative Writing | 83.1% | — |
| WildBench | 85.5% | — |
| EQ-Bench 4 | — | 1226 |
Frequently asked questions
Is GPT-5 Mini better than Inkling?
Inkling is the stronger model overall, scoring 44.1 to 41.8 on the Noometry Index. GPT-5 Mini costs 3.7× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, GPT-5 Mini or Inkling?
GPT-5 Mini is cheaper. It lists at $0.25 per million input tokens and $2 per million output tokens; Inkling lists at $1.87 and $4.68.
Is GPT-5 Mini or Inkling better for coding?
GPT-5 Mini scores higher on coding benchmarks: 40.1 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5 Mini does, with 400K tokens against 66K.
How many benchmarks do GPT-5 Mini and Inkling share?
34 benchmarks have published results for both models. GPT-5 Mini has 60 scored results on Noometry and Inkling has 41.