Remote Labor Index
Das Ergebnis jedes Modells, so, wie der Benchmarking Hub von Epoch AI es veröffentlicht, von Epoch gemessen oder aus der genannten Quelle gesammelt. Facthory hat sie nicht selbst gemessen.
Daten: Epoch AI, „Capabilities & benchmarking“, epoch.ai/benchmarks, lizenziert unter CC BY 4.0. Gelesen am 16.11.2026.
| # | Modell | Organisation | Ergebnis | Nachweis |
|---|---|---|---|---|
| 1 | Claude Fable 5.2 (unknown) API access | Anthropic | 41,3 % | – |
| 2 | Claude Opus 5.5 (unknown) API access | Anthropic | 21,3 % | – |
| 3 | GPT-6 Astra (unknown thinking) API access | OpenAI | 20,8 % | – |
| 4 | Claude Fable 5.1 (unknown) API access | Anthropic | 17,9 % | – |
| 5 | Gemini 4 Argon (unknown) API access | Google DeepMind | 17,1 % | – |
| 6 | Claude Fable 5 API access | Anthropic | 16,1 % | – |
| 7 | Claude Fable 5 (unknown) API access | Anthropic | 15,8 % | – |
| 8 | Claude Opus 4.8 API access | Anthropic | 8,3 % | – |
| 9 | GPT-5.5 (unknown thinking) API access | OpenAI | 6,3 % | – |
| 10 | Gemini 3.8 Flash (unknown) API access | Google DeepMind | 5,8 % | – |
| 11 | Gemini 3.7 Flash (unknown) API access | Google DeepMind | 5 % | – |
| 12 | Claude Opus 4.6 (unknown thinking) API access | Anthropic | 4,2 % | – |
| 13 | Claude Opus 4.5 (unknown thinking) API access | Anthropic | 3,8 % | – |
| 14 | GPT-5.2 (medium) API access | OpenAI | 2,5 % | – |
| 15 | Claude Sonnet 4.5 (unknown thinking) API access | Anthropic | 2,1 % | – |
| 16 | GPT-5.2 (unknown thinking) API access | OpenAI | 2,1 % | – |
| 17 | GPT-5 (unknown thinking) API access | OpenAI | 1,7 % | – |
| 18 | Gemini 3 Pro Preview API access | Google DeepMind | 1,3 % | – |
| 19 | Gemini 2.5 Pro Preview (Jun 2025) API access | Google DeepMind | 0,8 % | – |
