GPT-5.6 Sol
Alias for flagship general-purpose models with safeguards for defensive cybersecurity work
Shogo Agent Index
Supported#14 of 9261.8range 57.1–66.5 · 43% of weight measured
Agentic coding
71
54% measured
Long-horizon autonomy
72
50% measured
Knowledge-work agents
46
32% measured
Computer use
53
60% measured
Security & context learning
58
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 86 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 91 | |
| DeepSWE | 72.7% | 6% | mini-swe-agent · $8.39/task | |
| FrontierSWE | 32.2% | 5% | proximus · $179.64/task | |
| CursorBench | 41.7% | 4% | source · $8.23/task | |
| GSO-Bench | 76.5% | 4% | OpenHands | |
| METR Time Horizons | est. | 10% | predicted 83 | |
| Vending-Bench 2 | $9,619 | 6% | source | |
| PostTrainBench | 31.0% | 4% | Codex CLI | |
| APEX-Agents | 51.4% | 8% | source | |
| GDPval | est. | 7% | predicted 62 | |
| Remote Labor Index | est. | 5% | predicted 11 | |
| DeepResearch Bench | est. | 5% | predicted 51 | |
| OSWorld 2.0 | 27.3% | 6% | source | |
| OSWorld | est. | 4% | predicted 91 | |
| Cybench | est. | 5% | predicted 89 | |
| CL-bench | est. | 5% | predicted 27 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| ARC-AGI | Reasoning | 97.5% | source |
| ARC-AGI-2 | Reasoning | 92.5% | source |
| BALROG | Long-horizon | 60.0% | source |
| Chess Puzzles | Reasoning | 68.0% | source |
| CritPt | Reasoning | 32.3% | source |
| DTBench | Knowledge | 96.0% | source |
| FrontierCode | Coding | 47.5% | codex |
| FrontierMath Tiers 1-3 | Math | 89.1% | source |
| FrontierMath Tier 4 | Math | 82.9% | source |
| GeoBench | Multimodal | 73.0% | source |
| GPQA Diamond | Reasoning | 93.5% | source |
| LMCA | Knowledge | 59.2% | source |
| MirrorCode | Coding | 20.0% | source |
| OTIS Mock AIME | Math | 100.0% | source |
| ProofBench | Math | 83.0% | source |
| SciCode | Coding | 57.8% | source |
| SimpleBench | Reasoning | 71.7% | source |
| SimpleQA Verified | Knowledge | 69.7% | source |
| WebDev Arena | Coding | 1617 | source |
| WeirdML v3 | Coding | 15.0% | codex_cli |
API hosts
Live prices and uptime for every host serving GPT-5.6 Sol via OpenRouter.
| Host | Input | Output | Cache read | Agent $/M | Uptime 24h | Quant |
|---|---|---|---|---|---|---|
| OpenAI | $2.00 | $10.00 | $0.20 | $0.92 | 100.0% | — |
| Amazon Bedrock | $4.00 | $20.00 | $0.40 | $1.85 | 100.0% | — |
| Azure | $4.00 | $20.00 | $0.40 | $1.85 | 100.0% | — |
| OpenAI | $4.00 | $20.00 | $0.40 | $1.85 | 100.0% | — |
| OpenAI | $8.00 | $40.00 | $0.80 | $3.70 | 100.0% | — |