Total 1,098
In April 2024 NVIDIA agreed to acquire Run:ai, an Israeli maker of Kubernetes-based GPU orchestration software, reported at about $700 million. The European Commission approved the deal in December and it closed on 30 December 2024, with the software to be open-sourced.
On 24 April 2024 Snowflake announced Arctic, a large language model aimed squarely at enterprise work such as SQL generation, coding and instruction following, and released it with training recipes as well as weights.
On 23 April 2024 SenseTime released SenseNova 5.0 and benchmarked it against GPT-4 Turbo, and its shares surged more than 30% before trading was suspended.
On 23 April 2024 Microsoft introduced the Phi-3 family of small language models, starting with Phi-3-mini, released on Azure AI Studio, Hugging Face and Ollama.
On 22 April 2024 Apple researchers posted the OpenELM paper to arXiv and released eight small language models, with training logs, checkpoints, configurations and code for running the models on Apple devices.
On 18 April 2024 Meta released Llama 3 in 8B and 70B sizes, both pretrained and instruction-tuned. The models were pretrained on over 15 trillion tokens and made available through major cloud providers and Hugging Face under the Meta Llama 3 Community License.
On 17 April 2024, MiniMax released the abab 6.5 series, a family of mixture-of-experts language models reported to support a 200,000-token context. It extended the company's proprietary abab line, which began with abab1 in 2022.
On 17 April 2024 Mistral AI announced Mixtral 8x22B, a sparse mixture-of-experts language model with 141 billion total but only 39 billion active parameters, released under Apache 2.0 with a 64K context window.
In April 2024 Boston Dynamics retired its hydraulic Atlas and introduced a fully electric humanoid, promising more strength, dexterity and range of motion, with Hyundai's factories as the first testing ground.
On 4 April 2024 Cohere introduced Command R+, a retrieval-augmented model for enterprise workloads. It was first offered on Microsoft Azure, where Microsoft said it works with Azure AI Search and Azure AI Content Safety.
On 3 April 2024 Stability AI released Stable Audio 2.0, a model that generates stereo tracks of up to three minutes, adds audio-to-audio transformation, and was trained on more than 800,000 licensed files.
On 2 April 2024 China's cyberspace administration began publishing which generative AI services had completed filing, and required launched apps to display the model name and filing number they use.
On 29 March 2024 Langboat released Mengzi3, an open-weight language model family in 8B and 13B sizes under Apache-2.0, with base weights first and a chat version following later.
On 28 March 2024 AI21 Labs released Jamba, an open-weights model that combines Mamba state-space layers with Transformer attention. It has 52 billion total parameters and a 256k-token context.
On 28 March 2024 xAI announced Grok-1.5, with better reasoning and a 128,000-token context window. About two weeks later it previewed Grok-1.5V, its first multimodal model.
Databricks released DBRX, a mixture-of-experts model shipped in base and instruct versions, on 27 March 2024. The company framed it as competition for Mixtral and Grok-1 rather than closed frontier systems, and reported it ahead of LLaMA2-70B Chat on every benchmark it published.
On 23 March 2024 StepFun made its public debut in Shanghai, announcing Step-1, Step-1V and a preview of the trillion-parameter Step-2, plus two consumer apps.
On 23 March 2024, Emad Mostaque stepped down as CEO and board director of Stability AI; the company named interim co-CEOs and began searching for a permanent leader.
On 19 March 2024, Sakana AI posted a paper on using evolutionary search to merge open models without extra training data or compute, and released a 7B Japanese model that also handles mathematics as an experiment.
On 19 March 2024 Microsoft announced that Mustafa Suleyman and Karén Simonyan were joining to form Microsoft AI, which took over Copilot and the consumer AI products.
On 18 March 2024 NVIDIA announced Blackwell, the GPU platform succeeding Hopper, claiming it can run real-time generative AI on trillion-parameter models at up to 25 times lower cost and energy.
