Anthropic's Claude Fable 5 Challenges GPT 5.5 and Gemini 3.1 Pro With Stronger Coding and Reasoning Skills

Anthropic has unveiled Claude Fable 5, its newest AI model for general users, positioning it as a strong challenger to OpenAI’s GPT 5.5 and Google’s Gemini 3.1 Pro in the fast-moving artificial intelligence race.

The company says Claude Fable 5 is its most powerful publicly available model so far, delivering major improvements in coding, research, reasoning, vision tasks, and tool usage. Alongside it, Anthropic has also introduced Claude Mythos 5, a more advanced version that is currently limited to selected cybersecurity experts and research partners.

The latest launch highlights how competition among leading AI companies is becoming increasingly intense, with each firm racing to build smarter and more capable models.

According to benchmark results shared by Anthropic, Claude Fable 5 outperformed several rival AI models in key tests. In a coding benchmark known as SWE-Bench Pro, Claude Fable 5 scored 80.3 percent, significantly ahead of GPT 5.5’s 58.6 percent and Gemini 3.1 Pro’s 54.2 percent.

The model also showed strong performance in software engineering tasks. On the FrontierCode Diamond benchmark, Claude Fable 5 achieved 29.3 percent, compared with 5.7 percent for GPT 5.5. In knowledge and reasoning-related evaluations, it scored 1932 points, ahead of GPT 5.5’s 1769 and Gemini 3.1 Pro’s 1314.

Anthropic says both Fable 5 and Mythos 5 are built on the same underlying technology. However, Fable 5 includes additional safety protections designed to limit responses in sensitive areas such as cybersecurity, biology, chemistry and AI model replication.

If a user’s request enters one of these high-risk areas, Fable 5 may automatically switch the task to Anthropic’s older Claude Opus 4.8 model. The company says this happens in fewer than five percent of user sessions.

Meanwhile, Claude Mythos 5 has some of these safeguards removed and is being made available only to trusted partners through Anthropic’s Project Glasswing initiative, which is being conducted with support from the US government.

Claude Fable 5 is available immediately through Anthropic’s API and Claude platform. The company has priced the model at $10 per million input tokens and $50 per million output tokens.

For subscribers, Anthropic is offering access to Fable 5 at no additional cost until June 22. After that, users may need to purchase usage credits unless the company decides to extend the free access period based on available capacity.

With benchmark scores that place it ahead of GPT 5.5 and Gemini 3.1 Pro in several important categories, Claude Fable 5 marks Anthropic’s strongest attempt yet to challenge the biggest names in the AI industry.

Read More: Claude Mythos Arrives in India as Anthropic Expands Cybersecurity Initiative