An evaluation by NIST's CAISI says DeepSeek V4 Pro lags behind leading US AI models by about eight months and is the most capable Chinese AI model to date (NIST)
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- DeepSeek V4 Pro is rated by NIST’s CAISI as eight months behind leading U.S. models, yet remains China’s most capable AI to date.
- NIST CAISI published the evaluation, marking the first formal benchmark for a Chinese large‑language model.
- DeepClaude integrates DeepSeek V4 Pro in a Claude Code agent loop, delivering a 17× cost reduction according to Hacker News.
- Google is rolling out Gemini AI across Chrome to millions and launching a Mac app, expanding its consumer AI footprint.
- Microsoft reports a minimum three‑month delay in OpenAI services caused by a “4” issue, highlighting operational headwinds.
Why it matters: Chinese AI firms gain a benchmark but still trail US rivals; US providers keep a ~8‑month lead, preserving market share and pricing power for firms like OpenAI and Microsoft.
