Canva Launches First Image Generation Model, Canva Craft
Canva has introduced its first image generation model, Canva Craft, designed to compete with models from Google, OpenAI, and others. The model is now available on the platform and via API.
Canva has introduced its first image generation model, Canva Craft, designed to compete with models from Google, OpenAI, and others. The model is now available on the platform and via API.
Amazon Polly now offers more natural speech output with 14 new voices across 9 locales, including Hindi and Finnish, available globally.
Uber has launched a company-wide rollout of ChatGPT Work, making the agentic platform available to over 30,000 employees globally.
Anthropic introduces a new app that lets users delegate tasks to an AI agent, which operates in the background and requires user approval for sensitive actions.
Hugging Face has launched an open-source AI model that can interpret DNA sequences and answer questions about genetic effects, available for local laptop use.
Google Cloud is now the only major cloud provider offering Gemini, Claude, and OpenAI models on one platform, with three new OpenAI models available for enterprise customers.
AMD's Micro-World 2 enables real-time video generation with low-latency interaction, running on a single AMD Instinct™ MI350X GPU.
NVIDIA's new model enables interactive generative simulation of clinical surgery using real-time kinematic control and video data from procedures.
Hugging Face's new static embedding model outperforms existing multilingual and English-only models in retrieval tasks, with notable gains in multilingual performance.
NVIDIA has expanded its Nemotron Retriever collection with new models for embedding, reranking, and document extraction, aimed at improving agent efficiency and retrieval accuracy.
NVIDIA introduces Kumo Forecast, a time-series forecasting model that integrates pretrained representations with historical pattern retrieval for enhanced interpretability.
OpenAI claims GPT-6 achieves 87.4% test pass rate in 20 languages, up from 38% for GPT-5.6 Terra.