Search AI Products and News

Google introduces a feature letting users turn off visible watermarks when generating images, videos, and music. Revealed by Gemini lead Josh Woodward, it will roll out to models like Nano Banana, Omni, and Lyria, first on Gemini web and Google video products.....

OpenAI's new image model, mona-lisa-1, appeared on Arena with SynthID watermark. Tests show it greatly improves realism, reducing the 'plastic' look, with richer details. It excels at generating artistic paintings and infographics. Clues suggest its training may have been completed last year.....

xAI launches Imagine Image 2.0 as Grok's new quality mode, featuring controllable layouts, clear text, and visual consistency for creative use. The model understands complex instructions, plans text layouts in dense compositions, keeps small text sharp, and supports multi-turn editing while preserving user-specified visual features. API access planned.....

Meitu launches MeituHub, an AI imaging production line platform. It offers enterprises a full-chain solution from requirement understanding to workflow building and final delivery. Integrating full-stack imaging tech, it allows creating custom visual workflows via natural language and generates reusable web apps, addressing fragmented generative AI tools to build stable, efficient imaging production.....

Tongyi Qwen launched the Qwen-Image-3.0 Pro and Standard models, ranking first in China and second among mainstream models on the Arena text-to-image leaderboard. Supports 4.5k-token long prompts, 10px-level fine text rendering, and 12 languages, with the Pro version starting at $0.04 per image. The core highlight is the first integration of generation and editing into a single architecture, achieving unified multi-capability.

Alibaba's Qwen-Image-3.0 is launched, with Pro and Standard APIs available. Text-to-image starts from 0.18 yuan/image. It ranks first among Chinese models on Arena.ai, excelling in long text comprehension and detailed rendering, supporting up to 4.5k tokens for complex layouts like storyboards, exam papers, and 9-grid infographics.....

SenseNova U1.5-Lite-Preview, an open-source lightweight multimodal model with 8B MoT parameters and NEO-unify architecture, supports native 4K image generation, fine textures, accurate Chinese/English text rendering, and stable image editing. It enhances long-form natural language and structured visual instruction understanding for precise multimodal generation under complex demands.....

Tencent's AI Assistant Yuanbao upgrades its image understanding function, emphasizing the idea of 'take a photo when something happens'. This update enhances text recognition accuracy and detail perception, covering scenarios such as healthcare, daily maintenance, contract review, social communication, and content creation. It can accurately interpret medical reports, drug instructions, and more, providing users with instant image-based Q&A and decision-making references.

Google urgently suspended its new AI image generation feature on Google Earth within 48 hours of launch due to risks of creating misinformation. The tool integrated the Nano Banana2 model, allowing users to generate fictional scenes on real satellite and aerial imagery. BBC tests showed it could produce convincing but misleading content like 'Eiffel Tower collapsing.' Google pledged to enhance safety measures.....

xAI upgrades the video model Imagine Video 1.5 with new image reference, voice reference, text-only video generation, and native 1080p output, enhancing character consistency and scene control. Text-to-video and 1080p features are now available on all platforms of Grok Imagine. Image and voice references are first available to US SuperGrok Heavy/Plus users, with broader coverage planned for the future.

Google Earth's AI image generation, powered by Nano Banana2, blends satellite, aerial, and 3D maps. Web users zoom into an area, enter prompts, and visually transform real-world places directly on geographic data.....

Google has integrated the AI image generation model Nano Banana2 into the web version of Google Earth, making it available to users worldwide. This feature combines satellite imagery, aerial photos, and 3D topographic data, allowing users to input a scene description at any location to generate high-precision visualized images.

Midjourney acquires social astrology app Co-Star amid undisclosed terms. Co-Star, with ~4.3M monthly actives, blends AI and human input to generate birth charts and horoscopes. The move signals Midjourney's diversification beyond AI image/video generation.....

Midjourney V8.2 model released, enhancing aesthetic quality and personalization. It produces more creative, bold, refined images with fewer low-quality outputs. The personalization module precisely captures private aesthetic preferences, delivering results that better match individual styles after accumulating works.....

Microsoft unveiled two proprietary AI models: MAI-Image-2.5-Pro for high-precision imaging with up to 84% GPU cost reduction, and MAI-Voice-2-Flash tailored for high-concurrency voice interaction. Trained on clean, traceable data without third-party distillation, both are designed from the ground up for Microsoft product users.....

Alibaba released Qwen-Image-3.0, supporting up to 4.5K-token prompts to generate complex content like formulas, geometry, logic, and multi-layer UI. It natively renders 12 languages and 20+ fonts, with 4.5x longer text vs. prior version, advancing commercial creative tasks such as film storyboards and knowledge diagrams.....

MeituHub, unveiled at the 2026 Imaging Festival, is an AI imaging production line launching Aug 5, marking Meitu's shift from single-scene AI tools to a full-process customized imaging workflow provider. Unlike prior vertical products, it integrates core capabilities for end-to-end solutions.....

Google Search's AI Overview will add an image generation feature, allowing users to directly input text to generate images without relying on existing web images. The feature uses the "Nano Banana2Lite" model, prioritizing speed and cost over ultimate image quality. It is currently limited to English users and is expected to be launched in the coming weeks.

Step Star launched the Step Edge model series for phones and vehicles, including Basic, Audio, GUI, and Gen versions. These models process text, image, and audio locally, enabling screen understanding, speech recognition, interface control, and image generation. Tool invocation latency is as low as 0.1s. Simple, frequent, or weak-network tasks run fully on-device; complex reasoning is handled by cloud, achieving efficient device-cloud collaborati....

Meta launched an AI image generator called Muse Image, which allows users to modify others' photos by "@"ing public Instagram accounts. However, the feature lacked a notification mechanism, leading to strong opposition from users and agencies like CAA, resulting in a privacy controversy.