1. Start with the official public sources
If you need concrete public release information, begin with the DeepBrainz Hugging Face organization and then use DeepBrainz Labs for research guidance. The main corporate site explains how the release family fits the product system, but Hugging Face and Labs are the deeper sources for release-facing questions.
For support, include the exact release page, release name, or file you are reading. The answer can change depending on whether you mean a supported public release, a research checkpoint, a long-window variant, or a community build.
2. Supported and experimental are not the same thing
The R-series material explicitly separates supported releases, long-window variants, research checkpoints, and community quantizations. That matters because a product page, model release, research note, and experiment can have different support expectations.
- Supported releases are the safest starting point for serious evaluation.
- Experimental variants are useful for exploration, not automatic production assumptions.
- Research checkpoints help explain the direction, but they are not the same thing as recommended starting points.
- Community builds can be useful, but support and reliability expectations differ.
3. Current public R1 release family
The current public DeepBrainz-R1 release family highlights three recommended starting points: R1-4B, R1-2B, and R1-0.6B-v2. The public release pages explain which releases are production-oriented, which are research variants, and how long-context versions differ.
4. How R1 fits the product stack
DeepBrainz-R is the research initiative behind the public release family. Lexopedia AI is the current production surface for autonomous knowledge work. AgentFoundry is the governed engineering-agent execution layer. Labs explains the research context and evaluation side behind that stack.
5. What support can help with
Support can help you find the right official page, understand what is supported versus experimental, and choose the right DeepBrainz product or research page. Support is less useful if the question assumes private guarantees that are not present in the public release notes or official docs.