GPT-5 Mini
Small GPT-5 for responsive agents, coding help, and everyday automation
Shogo Agent Index
Estimated#69 of 9230.4range 21.4–39.3 · 22% of weight measured
Agentic coding
32
46% measured
Long-horizon autonomy
36
30% measured
Knowledge-work agents
26
estimated
Computer use
21
estimated
Security & context learning
33
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | 64.7% | 8% | source | |
| Terminal-Bench 2.0 | 34.8% | 8% | spoox-o-m | |
| DeepSWE | est. | 6% | predicted 26 | |
| FrontierSWE | est. | 5% | predicted 5 | |
| CursorBench | est. | 4% | predicted 27 | |
| GSO-Bench | est. | 4% | predicted 9 | |
| METR Time Horizons | est. | 10% | predicted 65 | |
| Vending-Bench 2 | $-31 | 6% | source | |
| PostTrainBench | est. | 4% | predicted 18 | |
| APEX-Agents | est. | 8% | predicted 26 | |
| GDPval | est. | 7% | predicted 31 | |
| Remote Labor Index | est. | 5% | predicted 1 | |
| DeepResearch Bench | est. | 5% | predicted 45 | |
| OSWorld 2.0 | est. | 6% | predicted 3 | |
| OSWorld | est. | 4% | predicted 47 | |
| Cybench | est. | 5% | predicted 49 | |
| CL-bench | est. | 5% | predicted 17 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| AlgoTune | Coding | 1.38× | source |
| ARC-AGI | Reasoning | 54.3% | source |
| ARC-AGI-2 | Reasoning | 4.4% | source |
| Chess Puzzles | Reasoning | 24.0% | source |
| CritPt | Reasoning | 0.0% | source |
| DTBench | Knowledge | 80.5% | source |
| Fiction.LiveBench | Long context | 69.4% | source |
| FrontierMath Tiers 1-3 | Math | 46.7% | source |
| FrontierMath Tier 4 | Math | 12.2% | source |
| GPQA Diamond | Reasoning | 75.0% | source |
| Humanity's Last Exam | Reasoning | 19.4% | source |
| LMCA | Knowledge | 34.2% | source |
| MATH Level 5 | Math | 97.8% | source |
| OTIS Mock AIME | Math | 86.7% | source |
| ProofBench | Math | 9.0% | source |
| SciCode | Coding | 39.2% | source |
| SimpleQA Verified | Knowledge | 21.6% | source |
| VPCT | Multimodal | 40.2% | source |