GPT-5.4 mini
Strong small GPT for coding subagents, quick tool use, and high-volume work
Shogo Agent Index
UnrankedNot ranked yet: 1 agentic benchmark measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.
Agentic coding
40
estimated
Long-horizon autonomy
46
estimated
Knowledge-work agents
28
20% measured
Computer use
26
estimated
Security & context learning
39
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 71 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 50 | |
| DeepSWE | est. | 6% | predicted 34 | |
| FrontierSWE | est. | 5% | predicted 7 | |
| CursorBench | est. | 4% | predicted 31 | |
| GSO-Bench | est. | 4% | predicted 15 | |
| METR Time Horizons | est. | 10% | predicted 69 | |
| Vending-Bench 2 | est. | 6% | predicted 25 | |
| PostTrainBench | est. | 4% | predicted 20 | |
| APEX-Agents | est. | 8% | predicted 32 | |
| GDPval | est. | 7% | predicted 37 | |
| Remote Labor Index | est. | 5% | predicted 2 | |
| DeepResearch Bench | 36.3% | 5% | source | |
| OSWorld 2.0 | est. | 6% | predicted 4 | |
| OSWorld | est. | 4% | predicted 59 | |
| Cybench | est. | 5% | predicted 60 | |
| CL-bench | est. | 5% | predicted 19 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| ARC-AGI | Reasoning | 63.7% | source |
| ARC-AGI-2 | Reasoning | 18.9% | source |
| Chess Puzzles | Reasoning | 23.0% | source |
| CritPt | Reasoning | 10.0% | source |
| DTBench | Knowledge | 80.0% | source |
| FrontierCode | Coding | 27.0% | codex |
| FrontierMath Tiers 1-3 | Math | 51.2% | source |
| FrontierMath Tier 4 | Math | 9.8% | source |
| GPQA Diamond | Reasoning | 86.9% | source |
| LMCA | Knowledge | 40.8% | source |
| OTIS Mock AIME | Math | 88.9% | source |
| ProofBench | Math | 21.0% | source |
| SciCode | Coding | 52.1% | source |
| SimpleQA Verified | Knowledge | 29.4% | source |
| WebDev Arena | Coding | 1397 | source |