Open Source LLMs Tools

DeepSeek's 4th generation flagship model. V4-Pro (1.6T MoE) and V4-Flash (284B MoE) featuring 1M context, native multimodality, and Hybrid Attention for extreme efficiency.

Phi-3, a family of open AI models developed by Microsoft. Phi-3 models are the most capable and cost-effective small language models (SLMs) available, outperforming models of the same size and next size up across a variety of language, reasoning, coding, and math benchmarks.

Mixtral 8x7B, a high-quality sparse mixture of experts model (SMoE) with open weights. Mixtral outperforms Llama 2 70B on most benchmarks with 6x faster inference. It matches or outperforms GPT3.5 on most standard benchmarks. paper：https://arxiv.org/pdf/2401.04088.pdf news：https://mistral.ai/news/mixtral-of-experts/

Llama3 is a large language model developed by Meta AI. It is the successor to Meta's Llama2 language model. Online test address:huggingface.co/Meta-Llama-3-70B-Instruct

A strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.

Gemma 4 is Google's latest open source large language model series, built on the Gemini architecture, offering improved performance, longer context windows, and better multilingual support.

DeepSeek's first-generation reasoning models, DeepSeek-R1-Zero and DeepSeek-R1. DeepSeek-R1-Zero, a model trained via large-scale reinforcement learning (RL) without supervised fine-tuning (SFT) as a preliminary step, demonstrated remarkable performance on reasoning.

Qwen3 is the large language model series developed by the Qwen team at Alibaba Cloud. Qwen3-2507 includes Instruct and Thinking variants with 235B-A22B, 30B-A3B, and 4B sizes, 256K long context, and 1M-token input support in some scenarios.

A large language model open sourced by xAI