Blog
Latest news, tutorials, and insights about AI
Model Releases
Mistral Large 2: The 123B Open-Weight Frontier Model
Mistral AI unveils Large 2 with 128K context, 123B params, and open weights. Benchmarking vs GPT-4o & Llama 3.1.
7 min read
M
Model Releases
Meta Llama 3.1: The 405B Open-Source Benchmark
Meta releases Llama 3.1, a 405B parameter model matching GPT-4 performance with 128K context. Developers can now deploy this milestone open-source model.
6 min read
Model Releases
Mistral NeMo 12B: The New Standard for Efficient Open-Source AI
Mistral AI and NVIDIA unveil NeMo, a 12B parameter model optimized for single-GPU deployment with 128K context and Apache 2.0 licensing.
7 min read
Model Releases
InternLM 2.5 Release: Open-Source Reasoning Powerhouse from Shanghai AI Lab
Shanghai AI Lab unleashes InternLM 2.5, a 20B parameter model challenging top-tier reasoning benchmarks with open weights and massive context support.
5 min read
Model Releases
Gemma 2 Release: Google's New Open-Source AI Model
Google DeepMind releases Gemma 2. 9B and 27B models outperform 2x larger competitors. Apache 2.0 license. Build better AI apps today.
6 min read
Model Releases
Anthropic Unveils Claude 3.5 Sonnet: The Coding Powerhouse
Released June 20, 2024, Claude 3.5 Sonnet marks a historic milestone, outperforming GPT-4o and Gemini 1.5 Pro while offering double the speed of Opus at a fraction of the cost.
5 min read
Model Releases
DeepSeek Coder V2: The Open-Source GPT-4 Turbo Rival
DeepSeek AI releases a 236B MoE coding model that rivals GPT-4 Turbo performance across 338 languages, marking a new era in open-source development.
8 min read
Model Releases
NVIDIA Unveils Nemotron-4 340B: The Open-Source Powerhouse for Synthetic Data
NVIDIA releases Nemotron-4 340B, a 340B parameter open-source model optimized for synthetic data generation with a permissive enterprise license.
6 min read
Model Releases
Qwen2 Release: The 72B Open-Source Challenger to Llama 3
Alibaba Cloud unleashes Qwen2, a 72B parameter powerhouse under Apache 2.0, rivaling top closed models with superior multilingual capabilities.
5 min read