
Loading page…
Loading page…
Company intelligence / Live profile
Track Hugging Face's role in the AI economy through Superpower Daily reporting on models, policy, and tools. This live company page updates automatically as new launches, partnerships, product moves, and market signals are published.
Latest coverage
Key developments
Nous says targeted training cut responses classified as propaganda from 22% to 0.9%. Its released evaluation makes the scoring rules inspectable—not politically neutral.
The downloadable model pairs a million-token context window with competitive company-reported coding scores. Its evaluation notes show why those scores are not a blanket win over Claude.
Connected coverage
Your company watchlist
The MIT-licensed checkpoint targets Blackwell GPUs. NVIDIA’s tests show modest score changes, but running it requires a modified version of vLLM.
model