Meta's latest collection of multimodal models.
Llama 3.1 is a new state-of-the-art model from Meta available in 8B, 70B and 405B parameter sizes.
New state of the art 70B model. Llama 3.3 70B offers similar performance compared to the Llama 3.1 405B model.
An open weights function calling model based on Llama 3, competitive with GPT-4o function calling capabilities.
LFM2 is a family of hybrid models designed for on-device deployment. LFM2-24B-A2B is the largest model in the family, scaling the architecture to 24 billion parameters while keeping inference efficient.
New AI model better Llama 4 than the one by Meta
The Llama 4 models are Meta's flagship LLMs. These models leverage a mixture-of-experts architecture to offer industry-leading performance in text understanding and generation.
The Llama 4 collection of models are natively multimodal AI models that enable text and multimodal experiences. These models leverage a mixture-of-experts architecture to offer industry-leading performance in text and image understanding.
Works, thanks ollama team for supporting Llama4!
unsloth's v2.0 dynamic quants of Llama-4-Scout-17B-16E-Instruct-GGUF, Q2_K_XL(2.71-bit)
1.42B open-weight Llama-architecture language and coding model with a custom 49k BPE tokenizer and 32K context window for fast CPU and mobile inference.
llama.31 using cladue sonnets prompting system
From-scratch THOX base model: 123M Llama-arch decoder pretrained by THOX (2,000 steps, AMD MI300X); 49.3% HumanEval / 44.6% MBPP zero-shot. Q4_K_M GGUF, edge-sized. A base/completion model, not a chat assistant. Apache-2.0.