tech
Fable 5 vs GPT 5.5: Anthropic's model dominated every benchmark, then the government pulled it
Fable 5 topped GPT 5.5 on every major benchmark but was pulled by the US government after three days, making GPT 5.5 the top model you can actually use.
TL;DR
- Fable 5, Anthropic's latest AI model, was available for only three days before being shut down by the US government.
- Fable 5 outperformed OpenAI's GPT 5.5 on multiple benchmarks, including SWE-Bench Pro and the Chatbot Arena leaderboard.
- The US government cited a jailbreak vulnerability as the reason for the shutdown, an assessment disputed by Anthropic.
- GPT 5.5 is now the most capable AI model accessible to the public, not due to its own improvements but because Fable 5 was removed.
- The shutdown forces developers to use less capable models, impacting workflows that rely on advanced AI performance.
- Pricing for GPT 5.5 is also more favorable for developers compared to Fable 5.