Model Releases
Jamba 1.5: The New Hybrid MoE Standard for Long Context
AI21 Labs launches Jamba 1.5, combining Mamba and Transformer architectures for unprecedented speed and 256K context support in an open-weight release.
6 min read
Latest news, tutorials, and insights about AI
AI21 Labs launches Jamba 1.5, combining Mamba and Transformer architectures for unprecedented speed and 256K context support in an open-weight release.
Microsoft unleashes Phi-3.5, a 4B MoE model with 128K context, optimized for edge devices and strong reasoning capabilities.
xAI launches Grok-2 on 2024-08-13, delivering competitive performance on X platform with advanced reasoning and multimodal capabilities.
Naver launches HyperCLOVA X, a 104B parameter LLM optimized for Korean with 100K context. Compare specs, pricing, and use cases for developers.
Black Forest Labs releases FLUX.1, a 12B parameter rectified flow transformer that rivals closed-source models in image quality and text fidelity.
Mistral AI unveils Large 2 with 128K context, 123B params, and open weights. Benchmarking vs GPT-4o & Llama 3.1.
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.
Mistral AI and NVIDIA unveil NeMo, a 12B parameter model optimized for single-GPU deployment with 128K context and Apache 2.0 licensing.
Shanghai AI Lab unleashes InternLM 2.5, a 20B parameter model challenging top-tier reasoning benchmarks with open weights and massive context support.