Meta

Llama 3 70B Instruct ReferenceMeta

State-of-the-art 70B parameter LLM for diverse language tasks and applications.

Llama 3 8B Instruct LiteMeta

Llama 3 8B Instruct Lite: Advanced, fast and cheapest one text generation model optimized for dialogue, emphasizing safety and helpfulness

Llama 3.3 70B VersatileMeta

Llama-3.3-70B-Versatile is Meta's advanced multilingual large language model, optimized for a wide range of natural language processing tasks. With 70 billion parameters, it offers high performance across various benchmarks while maintaining efficiency suitable for diverse applications.

Llama 4 ScoutMeta

Efficient multimodal AI with extensive context processing

Llama Guard 2 (8B)Meta

Llama-based 8B model for safe LLM content classification

LLama-3 Chat (8B)Meta

Meta Llama 3 8B Chat boosts productivity and creativity through task-oriented, assistant-like interactions.

Meta Muse ImageMeta

Meta Superintelligence Labs' first image generation model. Reasons through a prompt before rendering, checks its own draft, and renders legible text inside images. $0.013 per image.

Meta Muse Image EditMeta

The editing counterpart to Meta Muse Image: point it at an existing photo and describe the change. Understands the frame and scopes edits to just what you asked for. $0.013 per image.

Muse Glimmer 30BMeta

Muse Glimmer 30B is Meta's multimodal chat model with a 131K-token context window, combining vision, reasoning and tool use.

Muse Spark 1.2Meta

Muse Spark 1.2 is Meta's multimodal chat model with a 1M-token context window, combining vision, reasoning and tool use for long-context tasks.

Muse Spark 1.3Meta

Meta Muse Spark 1.3 is a multimodal reasoning model from Meta that is especially good at agentic and coding workflows, including longer-horizon work across multiple steps.