Meta reportedly plans to release open-source versions of its upcoming AI models, according to Axios, with sources suggesting this includes successors to the company's Llama series. The report mentions Llama 4 "Maverik" as a 400-billion parameter model that debuted in April, though this timeline and naming don't align with Meta's publicly announced release schedule for Llama 3.2, which launched in September.
If accurate, this represents Meta's continued commitment to open-source AI as a competitive strategy against closed models from OpenAI and Anthropic. Meta's Llama models have become the foundation for countless developer projects precisely because they're freely available for commercial use. A 400-billion parameter model would be significantly larger than Llama 3.1's 405B version, potentially offering substantial capability improvements while maintaining open access.
The sparse sourcing raises red flags about this report's reliability. No additional major outlets have corroborated the story, and the timeline details conflict with Meta's known release patterns. The "Maverik" name hasn't appeared in any of Meta's official communications about their model roadmap, and the April debut date doesn't match any known Llama releases.
For developers banking on open-source alternatives to proprietary models, this uncertainty matters. If Meta does release a 400B+ parameter open model, it could reshape the competitive landscape and provide a powerful free alternative to GPT-4 class models. But until Meta confirms these plans officially, treat this as speculation rather than roadmap fact.