Chinese AI labs have developed an efficiency moat that may define the AI market's development over the coming years.

WeatherNext is an AI-powered ensemble forecasting model for global weather prediction. It utilizes a novel Functional Generative Network architecture, which enables it to generate forecasts 8x faster and with resolution up to 1-hour.

Breakthroughs in understanding the Earth that previously required complex analytics and years of iteration are now made possible in a matter of minutes.

When does access to agents able to negotiate on your behalf improve market efficiency and equitable outcomes? When does it not?

Opus 4.5 costs 67% more than Sonnet. But Opus 4.5 used 76% fewer tokens to reach the same outcome.

Our most complex pages, which took 20+ prompts to recreate in other tools, only required 2 prompts in Claude Design.

GPT‑5.5 delivers this step up in intelligence without compromising on speed: larger, more capable models are often slower to serve, but GPT‑5.5 matches GPT‑5.4 per-token latency in real-world serving, while performing at a much higher level of intelligence.

GPT‑5.5 delivers this step up in intelligence without compromising on speed: larger, more capable models are often slower to serve, but GPT‑5.5 matches GPT‑5.4 per-token latency in real-world serving, while performing at a much higher level of intelligence.

In a 1-million-token context, V4-Pro uses only 27% of the computing power required by its previous model, V3.2, while cutting memory use to 10%.

In our internal experiments, Android CLI improved project and environment setup by reducing LLM token usage by more than 70%, and tasks were completed 3X faster than when agents attempted to navigate these tasks using only the standard toolsets.

Figma has close to 2,000 employees - not all working on product engineering of course. I really doubt Anthropic even needed 10 to build Claude Design.

Eight out of eight models detected Mythos's flagship FreeBSD exploit, including one with only 3.6 billion active parameters costing $0.11 per million tokens.

Four researchers and software engineers estimated that a skilled human engineer would take 2 to 17 weeks to reimplement gotree, as AI successfully did in this work.

Performance: dev-browser: 3m53s, $0.88, 100% success rate — beats MCP configs, Chrome extensions, 'browser skill' stacks.

60 秒四路数据源并行采集，输出图文交错的研报。

Where training a language model took 167 minutes on eight GPUs in 2020, it now takes under four minutes on equivalent modern hardware.

Meta says its rebuilt pretraining stack can reach equivalent capability with >10× less compute than Llama 4 Maverick

Brandon told the team on a Monday that OKRs were due Wednesday—a turnaround that would have been absurd without this agent.

we can reach the same capabilities with over an order of magnitude less compute than our previous model, Llama 4 Maverick.

Closed Loop + Finite Demand = Efficiency Plays. AI bookkeeping categorizes transactions, reconciles accounts, files returns. Deterministic rules applied to numbers.

E2B & E4B · A new level of intelligence for mobile and IoT devices

In 23 months, the same capability that needed 1.8 trillion parameters now fits in 4 billion parameters. A 450x compression.

By using SAM, the Alta team has been able to process more than 20 million images without incurring exorbitant costs, allowing them to focus on building the best possible product for their users.

the productivity and quality improvements are likely due to a switch in the business professionals’ time allocation: less time spent on cranking out initial draft text and more time spent polishing the final result.