Model comparison
GPT-5.4 mini vs Inkling
GPT-5.4 mini and Inkling score almost the same on the Noometry Index (45.0 vs 44.1), so choose on price, context window or the category you care about most.
Last verified . 36 shared benchmarks.
Summary
- They share 36 benchmarks with published results for both. GPT-5.4 mini scores higher in 3 categories and Inkling in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 mini leads 45.5 to 31.3.
- The biggest single-benchmark swing is WeirdML: 60.3% for GPT-5.4 mini and 32.3% for Inkling.
- GPT-5.4 mini is cheaper at $0.75 / $4.50 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- GPT-5.4 mini accepts more context: 400K tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 mini | Inkling | |
|---|---|---|
| Provider | OpenAI | Thinking Machines Lab |
| Noometry Index | 45.0 | 44.1 |
| Released | 2026-03-17 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 400K | 66K |
| Max output | 128K | 66K |
| Input $ / M tokens | $0.75 | $1.87 |
| Output $ / M tokens | $4.50 | $4.68 |
| Results tracked | 46 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 mini leads
GPT-5.4 mini: 45.2 (#72), Inkling: 34.5 (#234)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| FrontierCode | 27% | 14% |
| LMArena WebDev | 1397 | 1413 |
| SciCode | 49.9% | 47% |
| WeirdML | 60.3% | 32.3% |
| LMArena Coding | 1438 | 1464 |
| ALE-Bench | 1,189 | 946 |
| FrontierSWE | — | 4.1% |
Agentic & Tool Use Too close to call
GPT-5.4 mini: 29.9 (#81), Inkling: 29.6 (#85)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
| DeepResearch Bench | 36.3% | — |
Reasoning Inkling leads
GPT-5.4 mini: 30.4 (#85), Inkling: 40.4 (#56)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| ARC-AGI-2 | 18.9% | 36.5% |
| ARC-AGI-1 | 63.7% | 79.5% |
| CritPt | 10% | 5.4% |
| Chess Puzzles | 24% | 21% |
| LMArena Hard Prompts | 1424 | 1451 |
| DTBench | 80% | 87.5% |
| LMCA | 40.8% | 37.6% |
| Epoch Capabilities Index | 148.84 | 148.54 |
| SimpleBench | — | 50% |
| Kagi LLM Benchmark | 37.9% | — |
| NYT Connections (extended) | 61.8% | — |
| Thematic Generalization | 61.7% | — |
| Mystery Game Puzzles | 11% | — |
| ForecastBench | 57 | — |
Math GPT-5.4 mini leads
GPT-5.4 mini: 45.5 (#75), Inkling: 31.3 (#225)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 33.3% |
| FrontierMath Tier 4 | 9.8% | 4.9% |
| OTIS Mock AIME 2024-2025 | 88.9% | 88.9% |
| ProofBench | 21% | 0% |
| LMArena Math | 1419 | 1479 |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge Inkling leads
GPT-5.4 mini: 51.5 (#67), Inkling: 55.1 (#49)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| GPQA Diamond | 86.9% | 88.3% |
| SimpleQA Verified | 29.4% | 40.3% |
| LMArena Expert | 1435 | 1465 |
| Vectara Hallucination Rate | 5.5% | — |
Multimodal Not comparable
GPT-5.4 mini: 39.7 (#56), Inkling: —
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| LMArena Vision | 1245 | — |
Multilingual Inkling leads
GPT-5.4 mini: 51.9 (#96), Inkling: 54.0 (#52)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| LMArena Non-English | 1405 | 1434 |
| LMArena Chinese | 1446 | 1490 |
| LMArena French | 1440 | 1458 |
| LMArena German | 1409 | 1446 |
| LMArena Japanese | 1374 | 1429 |
| LMArena Korean | 1368 | 1404 |
| LMArena Russian | 1417 | 1429 |
| LMArena Spanish | 1405 | 1448 |
Instruction Following Inkling leads
GPT-5.4 mini: 74.1 (#102), Inkling: 75.1 (#71)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| LMArena Instruction Following | 1405 | 1426 |
Long Context Too close to call
GPT-5.4 mini: 43.0 (#112), Inkling: 43.8 (#86)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| LMArena Longer Query | 1407 | 1434 |
Writing & Preference Inkling leads
GPT-5.4 mini: 64.0 (#58), Inkling: 65.2 (#51)
| Benchmark | GPT-5.4 mini | Inkling |
|---|---|---|
| LMArena Text | 1412 | 1441 |
| LMArena Creative Writing | 1370 | 1387 |
| EQ-Bench Creative Writing | 1665 | 1611 |
| LMArena Multi-Turn | 1429 | 1436 |
| EQ-Bench 4 | — | 1226 |
Frequently asked questions
Is GPT-5.4 mini better than Inkling?
GPT-5.4 mini and Inkling score almost the same on the Noometry Index (45.0 vs 44.1), so choose on price, context window or the category you care about most.
Which is cheaper, GPT-5.4 mini or Inkling?
GPT-5.4 mini is cheaper. It lists at $0.75 per million input tokens and $4.50 per million output tokens; Inkling lists at $1.87 and $4.68.
Is GPT-5.4 mini or Inkling better for coding?
GPT-5.4 mini scores higher on coding benchmarks: 45.2 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 mini does, with 400K tokens against 66K.
How many benchmarks do GPT-5.4 mini and Inkling share?
36 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Inkling has 41.