🔴 Anthropic just launched Claude Fable 5.1 and Mythos 5.1
Three months after the original Fable 5 launch - and the 19-day government shutdown - Anthropic is back with a significantly upgraded version.
The benchmark jumps are massive. On Terminal-Bench coding: Fable 5 scored 42%. Fable 5.1 scored 55.8%. Mythos 5.1 hit 60.9% - beating GPT-5.6 Sol's 37.3% by a wide margin.
On Terminal-Bench Science: 52.6% vs the original's 24.7%.
Same price as Fable 5. Cache read costs cut by up to 45%. Knowledge cutoff updated to June 2026 - five months fresher than before.
The jailbreak that triggered the government shutdown? Gray Swan and two external firms attacked the new safety filters. Zero critical jailbreaks found.
Fable 5.1 is broadly available now. Mythos 5.1 remains restricted to vetted cybersecurity and life science organizations only.
🔗 https://anthropic.com/news/claude-fable-5-1-mythos-5-1
#AINews #Anthropic
Follow AI Fuels for daily AI discoveries 🚀
Three months after the original Fable 5 launch - and the 19-day government shutdown - Anthropic is back with a significantly upgraded version.
The benchmark jumps are massive. On Terminal-Bench coding: Fable 5 scored 42%. Fable 5.1 scored 55.8%. Mythos 5.1 hit 60.9% - beating GPT-5.6 Sol's 37.3% by a wide margin.
On Terminal-Bench Science: 52.6% vs the original's 24.7%.
Same price as Fable 5. Cache read costs cut by up to 45%. Knowledge cutoff updated to June 2026 - five months fresher than before.
The jailbreak that triggered the government shutdown? Gray Swan and two external firms attacked the new safety filters. Zero critical jailbreaks found.
Fable 5.1 is broadly available now. Mythos 5.1 remains restricted to vetted cybersecurity and life science organizations only.
🔗 https://anthropic.com/news/claude-fable-5-1-mythos-5-1
#AINews #Anthropic
Follow AI Fuels for daily AI discoveries 🚀