Daily AI News

Google AI Releases Gemini 3.5 Transcribe: A Speech-to-Text Model Reporting 2.6% Average WER Across 85+ Languages

Google released Gemini 3.5 Transcribe, a speech-to-text model offered as two endpoints: a streaming version with sub-second latency (no speaker diarization) and a batch version that retains diarization and word timestamps at half the cost. The model achieves an average 2.6% word error rate across 85+ languages.

Read source →

Gemini Omni 1.1 Flash lets you build with more control

Gemini Omni 1.1 Flash introduces a new suite of creative controls and generative video capabilities for developers, extending the Gemini model family with finer-grained steerability.

Read source →

Cohere Releases Parse 5 (parse-v5.0): A 2.3B Vision Language Model That Turns Enterprise Documents Into Markdown

Cohere launched Parse 5, a 2.3B‑parameter vision language model that converts PDFs, slides, and images into Markdown with HTML tables, bounding boxes, and image descriptions. It is available via API ($1.50 per 1,000 pages) or dedicated Model Vault instances.

Read source →

Google’s AI note-taking app now allows you to interact with books

Google's Gemini Notebook now features an "Expert Intelligence" capability that lets users pull content from purchased Google Play Books into the notebook for question answering, planning, and content generation.

Read source →

Harness tackles influx of agent-delivered code with Code Repository and AI Code Review

Harness Inc. launched Agent‑Ready Harness Code Repository and AI Code Review to manage the growing volume of code produced by AI coding agents, offering automated review and repository services for development teams.

Read source →

Adobe is adding more AI to Photoshop

Adobe rolled out an AI‑heavy update to Photoshop that introduces a new optional "AI Assisted Editor" toolbar consolidating tools such as prompt‑based image editing, background removal, and generative fills.

Read source →

Nvidia closes in on Hugging Face acquisition

Nvidia has reportedly agreed to buy Hugging Face, the popular open‑source AI hub, for $12.9 billion, a move that would combine Nvidia’s chip dominance with a leading repository for models, datasets, and community tools.

Read source →

Pentake’s blacklisting of Anthropic was unlawful, US judge rules

A U.S. federal judge ruled that the Trump administration’s February sanctions designating Anthropic as a supply‑chain risk were unlawful, potentially saving the AI firm billions in lost business and reputational harm.

Read source →

AWS, Nvidia Expand Partnership With 2 Million More GPUs

AWS and Nvidia announced an expansion of their partnership to include an additional 2 million GPUs, extending collaboration into CPUs, government AI infrastructure, and robotics.

Read source →

Hugging Face is selling a cute $399 open source duck robot, Microduck

Hugging Face began selling the Microduck, a $399 open‑source robot duck that users can program via reinforcement learning to learn new tricks and behaviors.

Read source →