# ExploitGym → Hermes Index

Calibration · https://benchgap.net/calibration/exploitgym/current/to/hermesindex/current · data retrieved 2026-11-18

Hermes Index is estimated from ExploitGym with a Michaelis–Menten curve fitted on 6 models measured on both: y = 0.8648·x / (0.18140 + x), R² = 0.87, cross-validated error 5.5 pp. It is used for 7 estimates.

| Estimated model | ExploitGym | Hermes Index | Source |
|---|---|---|---|
| [GPT-6 Cyber](https://benchgap.net/model/gpt-6-cyber.md) | 48.5% | 62.9% | estimated ± 5.5 pp, low confidence |
| [Claude Mythos 5.2](https://benchgap.net/model/claude-mythos-5-2.md) | 38.1% | 58.6% | estimated ± 5.5 pp, low confidence |
| [GPT-5.6 Cyber](https://benchgap.net/model/gpt-5-6-cyber.md) | 32.1% | 55.3% | estimated ± 5.5 pp, low confidence |
| [Claude Mythos 5.1](https://benchgap.net/model/claude-mythos-5-1.md) | 23.4% | 48.7% | estimated ± 5.5 pp, low confidence |
| [Claude Mythos Preview](https://benchgap.net/model/claude-mythos-preview.md) | 17.5% | 42.5% | estimated ± 5.5 pp, low confidence |
| [MiMo-V2.6-Flash](https://benchgap.net/model/mimo-v2-6-flash.md) | 6.0% | 21.5% | estimated ± 5.5 pp, low confidence |
| [Muse Spark 1.1](https://benchgap.net/model/muse-spark-1-1.md) | 0.8% | 3.7% | estimated ± 5.5 pp, low confidence |

## Fitted on

| Model | ExploitGym | Hermes Index |
|---|---|---|
| [Claude Opus 5.5](https://benchgap.net/model/claude-opus-5-5.md) | 33.3% | 63.3% |
| [GLM-5.3](https://benchgap.net/model/glm-5-3.md) | 15.0% | 39.2% |
| [GPT-6.1 Sol](https://benchgap.net/model/gpt-6-1-sol.md) | 35.1% | 57.1% |
| [GPT-6 Astra](https://benchgap.net/model/gpt-6-astra.md) | 42.4% | 56.3% |
| [GPT-6 Luna](https://benchgap.net/model/gpt-6-luna.md) | 11.6% | 33.9% |
| [GPT-6 Sol](https://benchgap.net/model/gpt-6-sol.md) | 22.1% | 44.1% |
