Google: Gemma 3n 2B (free)

Google's Gemma 3n E2B IT is a multimodal, instruction-tuned model developed by Google DeepMind. It's designed to operate efficiently at an effective parameter size of 2B while leveraging a 6B architecture. Based on the MatFormer architecture, it supports nested submodels and modular composition via the Mix-and-Match framework, making it highly adaptable. Gemma 3n models are optimized for low-resource deployment, offering a 32K context length (effective 8K on Multi AI) and strong multilingual and reasoning performance across common benchmarks. This variant is trained on a diverse corpus including code, math, web, and multimodal data, making it versatile for various tasks. It supports streaming capabilities for dynamic interactions. Access this powerful model for free on Multi AI. It's best suited for chat applications, providing a context window of 8K tokens and a max output of 4K tokens. Pricing is $0.00 per 1M input/output tokens, making it completely free to use.