Anthropic releases Claude Fable 5.1 with reduced pricing and updated safeguards
The new model family includes Fable 5.1 and Mythos 5.1, with Fable 5.1 promising stronger performance and lower costs for agentic tasks. Mythos 5.1 is available to Project Glasswing participants only.
1 source · cross-referenced
- Anthropic released Claude Fable 5.1 and Mythos 5.1, addressing customer feedback on pricing, data retention, and safeguards.
Anthropic announced the release of two new model versions: Claude Fable 5.1 and Mythos 5.1. The company states that Fable 5.1 delivers stronger performance than its predecessor while reducing costs. Typical usage is said to cost around 25 percent less, and complex agentic tasks may see up to a 45 percent reduction due to lower pricing on cached, previously processed data.
Anthropic also highlighted updates to safeguards in Fable 5.1, describing them as more precise and less likely to block basic biology-related queries compared to Fable 5. Software vulnerability identification is now permitted for Fable 5.1, though Anthropic notes that certain cybersecurity tasks, such as penetration testing and binary-based vulnerability scanning, will still be redirected to Opus models.
The company addressed data retention concerns by introducing Enterprise Frontier Safeguards, which store data on customer cloud servers to provide what Anthropic calls “complete privacy.” This feature is planned to roll out later in the fall.
Mythos 5.1 is launching exclusively within Project Glasswing, Anthropic’s program for early access participants. Anthropic claims that Mythos 5.1 with low reasoning scores performs comparably to its predecessor set to Max reasoning on benchmarks.
Early access users quoted in the announcement include Every CEO Dan Shipper and Box CEO Aaron Levie, who praised Fable 5.1’s coding performance, efficiency, and conversational quality. Fable 5.1 is available on all platforms, while Mythos 5.1 remains limited to Project Glasswing participants.
- Oct 13, 2026 · Google DeepMind — Blog
Google DeepMind releases Gemini 4 Argon with domain-leading performance and multi-domain benchmarks
Trust79 - Oct 9, 2026 · Hugging Face
Hugging Face publishes Falcon ASR, a 1.6B-parameter Arabic speech recognition model with Emirati focus
Trust84 - Oct 8, 2026 · OpenAI — News
OpenAI disrupts two AI-enabled influence operations using its models
Trust79