A new open-source code reasoning model called DeepCoder-14B-Preview has hit the market. Built atop DeepSeek-R1 and Qwen2.5 using reinforcement learning (RL), it aims to provide more flexibility by combining high-performance code generation with reasoning capabilities for real-world applications. Its performance is said to be comparable to OpenAI’s o3-mini, “but with a smaller footprint,” say its developers, the research-driven AI companies Together AI and Agentica. “We democratize the recipe for training a small model into a strong competitive coder,” explains Together AI. Continue reading Researchers Debut Preview of DeepCoder Reasoning Model

Popular website hosting platform WordPress.com has launched an AI website builder that can facilitate creation of a relatively simple website in just a few minutes. Free to anyone who registers for an account, the tool uses a chat-style interface, making website building “as simple as having a conversation,” according to the company, which says the new feature “takes your input and instantly creates a fully designed, content-ready WordPress website, complete with text, layouts, and images.” At the moment, the tool is not able to create a complex site that requires e-commerce or other integrations. Continue reading WordPress AI Website Builder a Free Solution for Basic Sites

Among the many tech advancements unveiled at Google Cloud Next include a major generative media upgrade to Vertex AI, Google Cloud’s managed AI development platform. The new Vertex AI Media Studio lets enterprise users generate complete videos from scratch using text prompts. Lyria, Google’s text-to-music model is now available on Vertex in private preview. Both are subject to an “allowlist.” Chirp 3 now creates custom voices with just 10 seconds of audio input, while Imagen 3 has gained improved abilities for reconstructing missing or damaged portions of an image. Continue reading Vertex AI Movie Studio Can Create Videos from Start to Score

Google has debuted a new accelerator chip, Ironwood, a tensor processing unit designed specifically for inference — the ability of AI to predict things. Ironwood will power Google Cloud’s AI Hypercomputer, which runs the company’s Gemini models and is gearing up for the next generation of artificial intelligence workloads. Google’s TPUs are similar to the accelerator GPUs sold by Nvidia, but unlike the GPUs they’re designed for AI and geared toward speeding neural network tasks and mathematical operations. Google says when deployed at scale Ironwood is more than 24 times more powerful than the world’s fastest supercomputer. Continue reading Google Ironwood TPU is Made for Inference and ‘Thinking’ AI

YouTube is rolling out a new tool called Music Assistant to its Creator Music marketplace. Music Assistant uses generative AI to automatically add royalty-free instrumental background music to long-form videos. Accessed via a dedicated tab in Creator Music, Music Assistant provides more control over things like the style, mood and instruments of the desired music, which is requested via text prompts. Creator Music is available in the U.S. to creators enrolled in the YouTube Partner Program. YouTube is also experimenting with a Shorts feature that automatically synchronizes audio for short-form video creators. Continue reading YouTube Brings Generative Music Assistant to Video Creation

Google has turned its Firebase backend-as-a-service (BaaS) platform into a full-stack AI workspace called Firebase Studio that builds custom apps in a browser-based environment. Available to anyone with a Google account during its preview phase, Google says Firebase Studio will be useful to beginners and pros alike, with Gemini-powered AI agents that can be used to automate the process of building, launching and monitoring mobile and web apps and related infrastructure. Firebase Studio “includes everything developers need to create and publish production-quality AI apps quickly, all in one place,” the company announced at Google Cloud Next 2025. Continue reading Google Firebase Now Full-Stack App Developer in a Browser

Google’s Gemini coding assistant has gained agentic capabilities, available as part of Gemini in Android Studio, a subscription service for businesses designed to make app development for the Android ecosystem easier and more secure. This agent-centric “AI-powered cloud for developers and operators” is designed to infuse AI into all stages of application development, laying the groundwork for more rapid software creation cycles. The service is available to those who subscribe to Gemini Code Assist Standard or Enterprise editions. The new offering was unveiled at the Google Cloud Next 2025 developer conference in Las Vegas. Continue reading Google Pushes Gemini in Android Studio for App Developers

A bipartisan Congressional group has reintroduced the NO FAKES Act, a bicameral bill that aims to protect creative rights by safeguarding artist voices and likenesses from unauthorized use by AI and other digital replication. The announcement culminated the Recording Academy’s Grammys on the Hill Advocacy Day with a Wednesday press conference led by Senators Marsha Blackburn (R-Tennessee) and Chris Coons (D-Delaware) and Representatives Madeleine Dean (D-Pennsylvania) and Maria Salazar (R-Florida). Simultaneously, the TAKE IT DOWN Act — which aims to tackle unauthorized images of people — moved with Senate support to near approval in the House. Continue reading Proposed Federal Bills Aim to Curb Unauthorized Replication

San Francisco-based AI startup Deep Cogito has released five AI models in preview, making them available under an open-source license agreement. The models come in sizes 3B, 8B, 14B, 32B and 70B, with plans to release 109B, 400B and 671B versions in the weeks and months ahead. As for the current models, “each outperforms the best available open models of the same size, including counterparts from Meta, DeepSeek and Alibaba, across most standard benchmarks,” Deep Cogito claims, noting that the 70B model in particular “outperforms the newly released Llama 4 109B MoE model.” Continue reading Deep Cogito Is Out of Stealth with Hybrid Reasoning Models

Amazon has updated its Nova model series, with Nova Reel 1.1 now able to generate AI videos of up to two minutes as well as gaining a new ‘multi-shot’ feature. Announced in December, Nova Reel marked Amazon’s initial foray into generative video. AWS developer advocate Elizabeth Fuentes says that Nova Reel accommodates user prompts of up to 4,000 characters that can generate a series of six-second shots for a sequence totaling two minutes. The company also introduced the Nova Sonic real-time voice model that supports third-party enterprise development. Continue reading AWS Updates Nova Reels and Adds Nova Sonic Voice Model

ByteDance is testing TikTok for Artists, a free feature performers can use to manage music and mesh with fans. Aimed at generating exposure and building engagement, the feature also provides analytics — from fan insights to performance metrics and trend reports. The Artists platform is now being tested in Australia, Indonesia, Japan, Korea and New Zealand. Though the U.S. deadline for ByteDance to remove TikTok from Chinese control was extended by 75 days (a second time) from April 5 to June 19, the social media app’s future here is more uncertain than ever. Continue reading ByteDance Tests TikTok for Artists with Its U.S. Fate in Limbo

Adobe introduced new versions of Premiere Pro and After Effects at the 2025 NAB Show. For the first time, Adobe’s Premiere Pro editing software gets Firefly generative AI powers. Generative Extend lets users elongate scenes by generating a few extra frames at 4K in the Premiere Pro app, which also gains New Media Intelligence to help find the right clip in seconds. There is also drag-and-drop color management and automatic caption translation to over 27 languages. After Effects gets a more powerful preview playback engine, new 3D motion design tools and HDR monitoring. Continue reading Adobe Upgrades Premiere Pro, After Effects with AI Features

Blackmagic Design has released DaVinci Resolve 20 in public beta, touting 100 AI-centric improvements that aim to save time in a workflow. Among them is AI IntelliScript, which automatically generates a timeline using the transcribed audio in media clips and matching it to the script. From that, it constructs “a timeline of the best selected takes, with any alternative takes placed on additional tracks for editor review,” Blackmagic says. Another, AI MultiCam SmartSwitch, compiles a camera angle timeline based on who is speaking. Meanwhile, AI Audio Assistant puts what Blackmagic calls a “professional audio mix” just a click away. Continue reading Blackmagic Adds AI-Powered Features to DaVinci Resolve 20

Generative AI program Midjourney has issued V7 in alpha, marking its first new model in almost a year. Notable updates include personalization turned on by default, which users must first set up — a process Midjourney says takes 5 minutes — and can then toggle on or off at any time. Another new flagship feature, Draft Mode, lets users render lower resolution images at “half the cost and 10 times the speed,” according to Midjourney, emphasizing “it’s so fast that we change the prompt bar to a ‘conversational mode’ when you’re using it on Web.” Draft Mode also supports voice prompts. Continue reading Midjourney Launches V7 Image Generator with Voice Prompts

Vimeo is launching Vimeo Streaming, a turnkey platform that allows businesses, creators, and just about anyone to create and monetize their own branded streaming service. The announcement was timed to the 2025 NAB Show. Vimeo suggests the new offering — described as “a technological leap forward for content and media businesses” — is the next generation of its existing over-the-top product that is already helping creators earn $429 million per month. The approach is similar to YouTube with more design flexibility and brand control, but living on the Web, outside the cozy confines of social media with its built-in audiences. Continue reading Vimeo Streaming Helps Creators Build Subscription Platforms