LaunchAI ModelsJune 24, 2026
AA-Briefcase benchmark shows frontier models take >20 min per agentic task
1 source
More stories today
- Hugging Face Uses Chinese Model in Self-Defense
- Claude replays all players in daily racing game Swervle
- How AI guardrails are impeding the work of offensive cybersecurity researchers
- Fugu-Ultra v1.1 beats Fable 5 by 7.9 points in coding and reasoning
- audio.cpp 0.4 adds Higgs Audio v3 TTS 4B and Fish Audio S2 Pro