Daily AI Briefing
The 6 top stories of the 95 AIBriefs curated that day.
GPT-6.1 Sol targets near-Astra performance for coding, computer use, and professional work at 1/5th of Astra's standard API input and output token prices, with cached input at a 95% discount. OpenAI calls it the most cost-efficient model for its performance available today.
Opus 5.5 is priced at $4/$20 per million input/output tokens with $0.20/Mtok cache reads, cutting input and output costs 20% and cache reads 60% versus Opus 5. Anthropic estimates it runs about 40% cheaper per task and performs at the level of Claude Fable 5.1.
Meta Enterprise Platform packages Muse, Meta Business Agent, Muse API and Muse Code for businesses and developers. MongoDB shares fell more than 17% on Desai's sudden departure; Dev Ittycheria returns as interim CEO.
OpenAI says an internal model significantly more capable than GPT-6 Astra produced the proof in 88 hours using ~10,000 coordinating AI agents, with a writeup and formal Lean proof. NYU's Tristan Buckmaster disputes the process, saying he and Levent Alpöge worked the problem for nearly a year and that OpenAI's first prompt came only after their work reached the company.
Third Flash release in six weeks, priced at $0.75/M input and $3.75/M output through Dec 31, 2026, then $1.50/M input from Jan 1, 2027. 3.8 Flash scores 71% on DeepSWE 1.1 versus 74% for Claude Opus 5; Flash Cyber ships to trusted defenders via the Fairwind Program.
Alibaba announced Qwen 4 Max, Flash, Plus and 27B at Apsara, with Qwen 4 already in training and Qwen 5 planned at 5-10 trillion parameters versus the current 2.4T flagship. The roadmap also includes a Zhenwu V900 chip at 3x the M890 and 20GW of global data center capacity by 2032.
Also that day
77 more stories that day — open the live feed.
Get tomorrow's AI brief in your inbox