# Gert Labs → OSWorld 2.1 Offline

Calibration · https://benchgap.net/calibration/gertlabs/current/to/osworld-2-1-offline/current · data retrieved 2026-11-22

OSWorld 2.1 Offline is estimated from Gert Labs with a Michaelis–Menten curve fitted on 10 models measured on both: y = 2.0000·x / (1.20027 + x), R² = 0.49, cross-validated error 10.1 pp. It is used for 25 estimates.

| Estimated model | Gert Labs | OSWorld 2.1 Offline | Source |
|---|---|---|---|
| [Ember-1](https://benchgap.net/model/ember-1.md) | 58.0% | 65.2% | estimated ± 10.1 pp, low confidence |
| [Muse Ember](https://benchgap.net/model/muse-ember.md) | 56.1% | 63.7% | estimated ± 10.1 pp, low confidence |
| [GPT-5.3 Codex](https://benchgap.net/model/gpt-5-3-codex.md) | 52.0% | 60.4% | estimated ± 10.1 pp, low confidence |
| [Mercury 3](https://benchgap.net/model/mercury-3.md) | 50.6% | 59.3% | estimated ± 10.1 pp, low confidence |
| [Claude Opus 4.6](https://benchgap.net/model/claude-opus-4-6.md) | 48.1% | 57.2% | estimated ± 10.1 pp, low confidence |
| [GPT-5.2](https://benchgap.net/model/gpt-5-2.md) | 47.7% | 56.8% | estimated ± 10.1 pp, low confidence |
| [Qwen 3.6 Max (preview)](https://benchgap.net/model/qwen3-6-max-preview.md) | 47.5% | 56.7% | estimated ± 10.1 pp, low confidence |
| [Gemini 3 Pro](https://benchgap.net/model/gemini-3-pro.md) | 47.4% | 56.6% | estimated ± 10.1 pp, low confidence |
| [Claude Sonnet 4.6](https://benchgap.net/model/claude-sonnet-4-6.md) | 47.3% | 56.6% | estimated ± 10.1 pp, low confidence |
| [Claude Opus 4.5](https://benchgap.net/model/claude-opus-4-5.md) | 46.0% | 55.4% | estimated ± 10.1 pp, low confidence |
| [Gemini 3.1 Pro](https://benchgap.net/model/gemini-3-1-pro.md) | 45.7% | 55.1% | estimated ± 10.1 pp, low confidence |
| [GPT-5.1](https://benchgap.net/model/gpt-5-1.md) | 39.9% | 49.9% | estimated ± 10.1 pp, low confidence |
| [Kimi K2.5](https://benchgap.net/model/kimi-k2-5.md) | 35.5% | 45.6% | estimated ± 10.1 pp, low confidence |
| [Qwen3.7 Plus](https://benchgap.net/model/qwen3-7-plus.md) | 35.3% | 45.4% | estimated ± 10.1 pp, low confidence |
| [Claude Sonnet 4.5](https://benchgap.net/model/claude-sonnet-4-5.md) | 34.9% | 45.0% | estimated ± 10.1 pp, low confidence |
| [Gemini 3 Flash](https://benchgap.net/model/gemini-3-flash.md) | 34.2% | 44.4% | estimated ± 10.1 pp, low confidence |
| [Muse Glimmer 30B](https://benchgap.net/model/muse-glimmer-30b.md) | 32.7% | 42.8% | estimated ± 10.1 pp, low confidence |
| [Gemini 2.5 Pro](https://benchgap.net/model/gemini-2-5-pro.md) | 32.6% | 42.7% | estimated ± 10.1 pp, low confidence |
| [Qwen3.6-27B](https://benchgap.net/model/qwen3-6-27b.md) | 32.4% | 42.5% | estimated ± 10.1 pp, low confidence |
| [Grok 4](https://benchgap.net/model/grok-4.md) | 31.6% | 41.7% | estimated ± 10.1 pp, low confidence |
| [Qwen3.6 Plus](https://benchgap.net/model/qwen3-6-plus.md) | 31.4% | 41.4% | estimated ± 10.1 pp, low confidence |
| [Grok 4.20 Multi-agent](https://benchgap.net/model/grok-4-20-multi-agent-beta.md) | 30.2% | 40.2% | estimated ± 10.1 pp, low confidence |
| [MiMo-V2.5](https://benchgap.net/model/mimo-v2-5.md) | 29.9% | 39.9% | estimated ± 10.1 pp, low confidence |
| [Grok 4.1 Fast](https://benchgap.net/model/grok-4-1-fast.md) | 29.3% | 39.2% | estimated ± 10.1 pp, low confidence |
| [Qwen3.5 397B](https://benchgap.net/model/qwen3-5-397b.md) | 29.2% | 39.2% | estimated ± 10.1 pp, low confidence |

## Fitted on

| Model | Gert Labs | OSWorld 2.1 Offline |
|---|---|---|
| [Claude Fable 5.1](https://benchgap.net/model/claude-fable-5-1.md) | 77.0% | 83.5% |
| [Claude Fable 5.2](https://benchgap.net/model/claude-fable-5-2.md) | 88.2% | 89.3% |
| [Claude Haiku 5.5](https://benchgap.net/model/claude-haiku-5-5.md) | 58.6% | 72.4% |
| [Claude Opus 5.5](https://benchgap.net/model/claude-opus-5-5.md) | 83.3% | 87.2% |
| [Claude Sonnet 5.5](https://benchgap.net/model/claude-sonnet-5-5.md) | 69.9% | 83.9% |
| [Gemini 4 Argon](https://benchgap.net/model/gemini-4-argon.md) | 80.6% | 74.7% |
| [Gemini 4 Carbon](https://benchgap.net/model/gemini-4-carbon.md) | 83.7% | 81.5% |
| [Gemini 4 Hydrogen](https://benchgap.net/model/gemini-4-hydrogen.md) | 79.1% | 77.6% |
| [GPT-6.1 Terra](https://benchgap.net/model/gpt-6-1-terra.md) | 83.2% | 74.2% |
| [GPT-6 Luna](https://benchgap.net/model/gpt-6-luna.md) | 58.8% | 48.9% |
