High-performance text-generation model with 7,770,334 monthly downloads
Qwen3 4B (Qwen/Qwen3-4B) is an industry-leading open-weights model hosted on Hugging Face. Supporting state-of-the-art inference across modern AI workloads.
Compare pricing models, feature tradeoffs, and zero-cost open-source replacements.
Frontier reasoning and code generation models at open-source efficiency
High-performance image-text-to-text model with 7,153,238 monthly downloads
High-performance text-to-image model with 742,883 monthly downloads
High-performance text-generation model with 778,413 monthly downloads
Conversational AI agent capable of writing, reasoning, and analysis
Get up and running with Llama 3, DeepSeek, and Mistral locally