Releases

DeepSeek LLM 67B Base became publicly available as part of DeepSeek's first general LLM release.

DeepSeek published its first general-purpose 7B and 67B open language models.

DeepSeek LLM 7B Base became publicly available as part of DeepSeek's first general LLM release.

DeepSeek LLM 67B Chat became publicly available as part of DeepSeek's first general LLM release.

DeepSeek LLM 7B Chat became publicly available as part of DeepSeek's first general LLM release.

Perplexity introduced PPLX 7B Online and PPLX 70B Online, language models grounded with current web search.

Amazon released a text-to-image generation and editing model through Bedrock.

Amazon released a multimodal embedding model mapping text and images into a shared space.

Google DeepMind reports that GNoME identified 2.2 million new crystal structures, including hundreds of thousands of stable candidates.

OpenAI announced Sam Altman's return as CEO and an initial board chaired by Bret Taylor.

AWS and NVIDIA expanded their partnership around GPU infrastructure, DGX Cloud, NVIDIA software and future AI supercomputing.

AWS announced its second-generation AI training accelerator with substantially higher performance.

AWS introduced resilient infrastructure for training large foundation models across thousands of accelerators.

AWS announced real-time generative-AI assistance for contact-center agents.

AWS announced configurable safeguards for foundation-model inputs and outputs.

AWS introduced an assistant for developing on AWS, troubleshooting and transforming application code.

AWS announced a new family of generative-AI assistants for work and software development.

AWS introduced an enterprise assistant that answers questions and completes tasks using company data.

AWS introduced automatic and human evaluation workflows for comparing foundation models.

Baidu fully opened ERNIE APIs to enterprise customers through AI Cloud.

Anthropic released Claude 2.1 with a 200K-token context window, lower hallucination rates, system prompts, and beta tool use.

Microsoft Research releases Orca 2 in 7B and 13B sizes, training smaller models to use different reasoning strategies.

OpenAI's board announced that Sam Altman would depart as CEO and Mira Murati would serve as interim CEO.

AWS launched a free public playground for building shareable generative-AI applications with Amazon Bedrock.

Meta published Emu Edit for instruction-based image editing with precise task control.

Meta published Emu Video, a factorized text-to-video generation method.

Google DeepMind and YouTube introduce Lyria alongside experimental music-generation tools Dream Track and Music AI Tools.

Databricks introduced a data intelligence engine that learns enterprise semantics and powers context-aware assistance and optimization.

xAI began early API access for Grok developers ahead of the later public beta.

Microsoft introduces Maia 100, its first custom AI accelerator for Azure training and inference workloads.

Microsoft introduces Copilot for Azure to help customers design, operate and troubleshoot cloud infrastructure using natural language.

Microsoft launches Copilot Studio as the low-code platform for customizing Microsoft Copilot and building standalone copilots.

Microsoft renames Bing Chat and Bing Chat Enterprise to Microsoft Copilot and expands commercial data protection.

Google DeepMind publishes GraphCast, a machine-learning model for fast, accurate medium-range global weather forecasts.

CoreWeave announced plans to offer NVIDIA GH200 Grace Hopper Superchip instances in the first quarter of 2024 and deploy NVIDIA HGX H200 systems.

NVIDIA introduced H200, adding HBM3e memory to the Hopper platform for generative AI and HPC.

Scale launched its Safety, Evaluations, and Alignment Lab to build standardized frontier-model benchmarks, evaluation products, and red-team methods.

xAI launched PromptIDE for prompt engineering and interpretability research on Grok-1.

OpenAI introduced user-configurable GPTs for creating tailored versions of ChatGPT.

OpenAI launched the Assistants API beta with Code Interpreter, Retrieval, and function calling.

OpenAI announced GPT-4 Turbo with a 128K context window, newer knowledge, and lower API pricing.

xAI announced Grok and its first public model after two months of training, with real-time knowledge through X.

Base and instruction-tuned DeepSeek Coder checkpoints were published for code completion and generation.

DeepSeek released its first open code-model family, spanning 1B to 33B parameters and trained on two trillion tokens.

Cohere introduced its third-generation embedding models with stronger multilingual retrieval and compression options.

Moonshot exposed a 128K-context text model through its API.

Moonshot exposed a 32K-context text model through its API.

Moonshot introduced developer API access to its language models.

Moonshot exposed an 8K-context text model through its API.

Baidu began charging a subscription for public access to ERNIE 4.0 through ERNIE Bot.

DeepSeek launched its web conversational assistant.

DeepSeek began offering web-based access to its models through the DeepSeek Chat service.

Cohere released third-generation English and multilingual reranking models for enterprise search.

Pollen Robotics launched Reachy 2, an open-source humanoid robot platform later brought into Hugging Face through its acquisition of Pollen.

Hugging Face released a production-grade server for serving text embedding and reranking models.

Microsoft 365 Copilot becomes generally available to enterprise customers across productivity applications and Microsoft 365 Chat.

Google agrees to invest up to $2 billion in Anthropic, deepening a strategic AI partnership that also includes Google Cloud infrastructure.

Google agreed to invest up to $2 billion in Anthropic, including $500 million upfront and an additional $1.5 billion over time.

IBM released generative coding assistants for Red Hat Ansible automation and mainframe application modernization.

Hugging Face H4 released an improved Zephyr 7B beta model trained on expanded public and synthetic preference data.

Meta introduced ExecuTorch, an open-source framework for efficient on-device PyTorch inference.

Databricks acquired enterprise data replication company Arcion to make operational data available for analytics and AI.

IBM Research published a memory-centric neural inference chip designed for high energy efficiency.

NVIDIA publicly released its open-source TensorRT-LLM library for optimized large-model inference.

OpenAI made DALL-E 3 available to ChatGPT Plus and Enterprise users.

Baidu launched ERNIE 4.0 with upgraded comprehension, reasoning, generation and memory.

Microsoft makes Azure AI Content Safety generally available for detecting harmful content in text and images.

Scheduled and cron-triggered functions became generally available.

HTTP endpoints backed by autoscaling Modal functions became generally available.

Modal's programmatic container image system became generally available.

Versioned Modal application deployments became generally available.

Modal's Python-first SDK and developer workflow reached general availability.

Serverless Modal functions became generally available with elastic scaling.

Modal made serverless model deployment and inference generally available.

Modal made autoscaling CPU and GPU compute generally available as part of its cloud launch.

Modal announced a $16 million Series A, bringing total capital raised to $23 million.

Modal launched its serverless cloud platform into general availability.

The web-based Kimi conversational assistant became publicly available with long-context document capabilities.

Moonshot launched Kimi as a long-context AI assistant for the Chinese market.

Perplexity launched pplx-api with hosted access to its PPLX chat models.

Perplexity released open-weight PPLX 7B Chat and PPLX 70B Chat models.

Microsoft Research releases AutoGen, a framework for building applications with conversing and collaborating AI agents.

IBM made its first 13B Granite enterprise language models available through watsonx.ai.

IBM released the first enterprise-focused Granite language and code models in watsonx.

Hugging Face H4 released an aligned 7B chat model trained with distilled direct preference optimization.

Amazon released Titan Text Express for general text-generation workloads through Bedrock.

Amazon released its lower-cost Titan text-generation model through Bedrock.

AWS launched managed retrieval-augmented generation for grounding foundation models in private data.

AWS made Amazon Bedrock generally available with managed access to Amazon and third-party foundation models.

Microsoft and Nuance introduce DAX Copilot, combining ambient clinical intelligence with generative AI for automated documentation.

Claude became available to all Amazon Bedrock customers through the Anthropic-AWS partnership.

Databricks introduced pay-per-token APIs and provisioned throughput for serving curated open foundation models.

Mistral AI released Mistral 7B, its first open-weight language model, under the Apache 2.0 license.

Meta and EssilorLuxottica launched the second-generation Ray-Ban Meta smart glasses with voice and AI capabilities.

Meta introduced Emu as the image-generation technology powering new generative AI experiences.

Meta introduced Meta AI, a conversational assistant for WhatsApp, Messenger and Instagram.

Microsoft begins rolling out Copilot in Windows as part of a Windows 11 feature update.

Amazon agreed to invest up to $4 billion in Anthropic, with AWS becoming Anthropic's primary cloud provider and Trainium and Inferentia used for model development.

Amazon made an initial $1.25 billion investment in Anthropic under its broader commitment of up to $4 billion.

Amazon agreed to invest up to $4 billion and make AWS Anthropic's primary cloud provider for mission-critical workloads.