Google is rolling out Gemini Go, a streamlined version of its Gemini AI assistant designed for Android Go devices with 2GB of RAM or more. The update replaces Assistant Go and integrates into the Google Search app on entry-level Android phones.
Google DeepMind has released Gemma 4 12B Unified, an encoder-free multimodal model that processes text, images, and audio through a single decoder-only transformer. The model features 11.95 billion parameters, a 256K token context window, and achieves 77.2% on MMLU Pro and 72.0% on LiveCodeBench v6.
Replit has launched SEO Agent, an automated tool that scans deployed applications for search optimization issues and can automatically apply fixes. The agent checks crawlability, meta tags, semantic structure, and generates recommendations ranked by impact, with changes applied as separate reviewable tasks.
Meta has launched AI agents for businesses on WhatsApp, Instagram, and Messenger that can close sales, recommend products, and book appointments. Over one million businesses in India, Mexico, and Brazil have already signed up during testing. The feature is currently free but will move to a paid subscription in the coming months.
Meta is launching Meta Business Agent, an AI service that businesses can use across WhatsApp, Messenger, and Instagram to respond to customer inquiries, recommend products, and book appointments. The feature will be part of the Meta One subscription tier announced last week, marking Meta's latest attempt to diversify beyond advertising, which still accounts for 98% of its revenue.
Alibaba's Qwen has released Qwen3.7 Plus, a multimodal model with a 1 million token context window. The model accepts text and image input with text output, priced at $0.40 per million input tokens and $1.60 per million output tokens through OpenRouter's API.
ByteDance released Bernini-R, an open-source video generation and editing model that combines an MLLM-based semantic planner with a DiT-based renderer. The model requires Hopper-class GPUs (H100/H800/H200) for optimal performance and supports multiple tasks including text-to-video, video editing, and reference-guided generation.
Microsoft announced two new language models: MAI-Thinking-1, a 35B parameter reasoning model available to select early partners, and MAI-Code-1-Flash, a 5B parameter coding model rolling out to GitHub Copilot individual users in VS Code. Both models were trained on commercially licensed data without distillation from third-party models.
OpenRouter has released Fusion, a multi-model routing system that processes prompts through parallel expert model panels with web search enabled, then uses a judge model to synthesize consensus, contradictions, and unique insights. Users pay the sum of all underlying model completions rather than a single model price.
Microsoft released ASSERT (Adaptive Spec-driven Scoring for Evaluation and Regression Testing), an open-source framework that converts natural language descriptions of expected AI behavior into structured test cases. The tool addresses a gap in AI evaluation by testing application-specific behaviors that general benchmarks cannot capture.
Microsoft released seven AI models at Build 2026, headlined by MAI-Thinking-1, its first reasoning model with 35 billion parameters. The company claims the model matches Anthropic's Claude Opus 4.6 on SWE Bench Pro coding benchmarks and beats Sonnet 4.61 in blind tests.
Google is rolling out interface updates to the Gemini overlay on Android. The overlay now supports Dynamic Color theming in version 17.28 of the Google app, with the Screen content capture feature relocated from the tools menu to the main carousel alongside Photos, Camera, Files, Drive, and Notebooks.
Microsoft announced two proprietary AI models at its Build developer conference: MAI-Code-1 for code generation and MAI-Thinking-1 for reasoning tasks. The models are designed to run on Azure infrastructure, allowing Microsoft to reduce costs from its $13 billion OpenAI investment while competing directly with Anthropic and Google.
Microsoft announced MAI-Thinking-1, its first advanced reasoning AI model, at Build 2026. The company claims it's a medium-sized model matching leading models on key software engineering benchmarks, trained from scratch without distillation from third-party models.
Microsoft has launched Scout, an AI assistant built on the OpenClaw framework that operates across Microsoft 365. The system requires a GitHub Copilot subscription and includes policy conformance checks with audit trails to address security concerns about autonomous AI agents.
Perplexity announced that its Computer agentic system will gain hybrid inference in July 2026, automatically splitting tasks between local models for sensitive data and cloud-based frontier models for complex operations. The feature aims to balance privacy with computational power without requiring manual model selection.
NVIDIA released Cosmos3-Super-Text2Image, a 64-billion parameter text-to-image generation model as part of its Cosmos3 collection of omnimodal world models. The model uses a Mixture-of-Transformers architecture combining autoregressive and diffusion transformers, designed for Physical AI applications including robotics and autonomous vehicles.
AWS has published a technical guide on hyperparameter optimization for Amazon Nova Forge, its platform for building custom frontier models from Amazon Nova checkpoints. The guide addresses three core challenges: catastrophic forgetting during domain specialization, learning rate calibration when mixing proprietary and curated training data, and baseline performance constraints for reinforcement fine-tuning.
AWS published a technical guide showing how Amazon Nova 2 Lite performs object detection through natural language prompts without requiring model training. The multimodal model returns bounding box coordinates in JSON format at $0.0003 per thousand input tokens and $0.0025 per thousand output tokens, with typical images costing approximately $0.00057 to process.
GitHub announced a new Copilot desktop application at Microsoft Build 2026. The standalone app introduces what the company calls an "agent-native" interface designed to integrate AI agents into developer workflows.