Meta Hands Over a 405-Billion-Parameter Model
Meta open-sourced Llama 3.1 405B under a permissive license. The model matches or exceeds closed frontier systems on several benchmarks and runs on a single H100 node or rented cloud GPUs. Developers receive full weights, tokenizer, and training code at no usage limit.
Teams no longer need to pay per-token API fees or accept rate limits. You can fine-tune the model on private data and deploy it on hardware you control. The change replaces cloud dependency with local ownership of both inference and training.
Meta released the weights on Hugging Face and reports thousands of downloads in the first week. Independent labs have already fine-tuned the 405B model on domain-specific datasets and achieved parity with GPT-4 on internal evaluations.
Step 1: Visit https://ai.meta.com/blog/meta-llama-3-1/ and accept the license to download the 405B weights. Step 2: Load the model with Hugging Face Transformers on an 8xH100 instance or equivalent rented GPUs. Step 3: Run inference locally or begin supervised fine-tuning; expect no token caps and full control over the weights.