1,497Anthropic's Managed Agents Gets Budget Caps, Geo-Pinning and Smarter Advisor ModelsClaudeDevs·Agents·3 hrs ago·
32,668Anthropic's Claude Code Lets AI Sessions Talk Directly Without Human MiddlemenClaudeDevs·Development·6 hrs ago·
6,818OpenAI's Astra Becomes First AI Flagged as Critically Dangerous Before ReleaseOpenAI·Security·7 hrs ago·
10,905Anthropic's Claude Code Auto Mode Catches Dangerous Commands 89% of the TimeClaudeDevs·Development·7 hrs ago·
921Prime Intellect Ships Multi-Agent RL Training Into Its Open-Source StackPrime Intellect·Agents·8 hrs ago·
201Artificial Analysis Rebuilds Its Image Arena to Rank Models by Real WorkArtificial Analysis·Benchmarks·9 hrs ago·
1,379MiniMax Rebuilds Code 2.0 on Pi Agent, Slashing Latency by 90%MiniMax_Agent·Development·15 hrs ago·
1,426OpenAI's Codex Security Review Catches 74% of Real Bugs Semgrep MissesOpenAI Developers·Development·1 day ago·
347Google's Gemini 3.6 Flash Hits 91.2% on the World's Hardest Reasoning TestARC Prize·Benchmarks·1 day ago·
