Breaking

Google DeepMind releases Gemma 4 with multimodal reasoning and up to 256K context window

Google DeepMind released Gemma 4, a multimodal model family supporting text, images, video, and audio with context windows up to 256K tokens. The release includes four sizes (E2B, E4B, 26B A4B, and 31B) designed for deployment from mobile devices to servers. The 31B dense model achieves 85.2% on MMLU Pro and 89.2% on AIME 2026.

April 4, 2026

Latest News

All news →
0
product updateAnthropic

Anthropic blocks Claude subscriptions from OpenClaw access, requires separate pay-as-you-go billing

Anthropic is effectively blocking Claude subscription access to third-party tools like OpenClaw starting April 4, 2026 at 3PM ET. Users will need to purchase separate pay-as-you-go usage bundles to continue using OpenClaw with Claude. The move comes as OpenClaw's popularity has strained Anthropic's infrastructure capacity.

2 min readvia theverge.com
0
model releaseDeepSeek

Deepseek v4 launching on Huawei chips exclusively, signaling China's AI independence progress

Deepseek v4 is launching in the coming weeks running exclusively on Huawei chips, marking a major milestone in China's effort to reduce dependency on foreign semiconductors. Chinese tech giants including Alibaba, Bytedance, and Tencent have ordered hundreds of thousands of Huawei Ascend 950PR units to deploy the model through their cloud services.

2 min readvia the-decoder.com
0
analysis

Gemma 4 success hinges on tooling and fine-tuning ease, not benchmark scores

Google's Gemma 4 release marks a shift in open model strategy with Apache 2.0 licensing and competitive benchmarks, but real success depends on factors rarely measured: tooling stability, fine-tuning ease, and ecosystem adoption. The open model landscape is now crowded with alternatives like Qwen 3.5, Nemotron 3, and others—a maturation that changes what separates winners from the field.

0
product updateAnthropic

Anthropic attributes Claude Code usage drain to peak-hour caps and large context windows

Anthropic has identified two primary causes for Claude Code users hitting usage limits faster than expected: stricter rate limiting during peak hours and sessions with context windows exceeding 1 million tokens. The company also recommends switching to Sonnet 4.6 instead of Opus, which consumes limits roughly twice as fast.

0
product updateOpenAI

OpenAI shifts Codex to usage-based pricing, offers $500 credits to enterprise customers

OpenAI is replacing per-seat licensing with usage-based pricing for Codex in ChatGPT Business and Enterprise plans, eliminating upfront license costs. Eligible Business customers can claim up to $500 in promotional credit per workspace. The shift targets enterprises where coding tools typically expand from individual developers to full teams, positioning OpenAI against GitHub Copilot and Cursor.

1 min readvia the-decoder.com
1
product updateOpenAI

ChatGPT now integrates with Apple CarPlay for hands-free conversation

OpenAI's ChatGPT is now available directly on Apple CarPlay, allowing drivers to conduct full voice conversations with the AI assistant while driving hands-free. The integration requires iOS 26.4, the latest ChatGPT app, and a compatible vehicle. Unlike Siri, ChatGPT cannot access device functions like email, messaging, or Maps, but provides information on complex topics Siri struggles with.

2 min readvia zdnet.com
0
model releaseZhipu AI

Zhipu AI releases GLM-5V-Turbo: multimodal model generates front-end code from design mockups

Zhipu AI released GLM-5V-Turbo, a multimodal coding model that converts design mockups directly into executable front-end code. The model processes images, video, and text with a 200,000-token context window and 128,000-token max output, priced at $1.20 per million input tokens and $4 per million output tokens.

1
product update

Cursor 3 rebuilds IDE around parallel AI agent fleets, moves away from classic editor layout

Cursor released version 3 of its AI coding tool with a complete interface redesign built around running multiple AI agents in parallel rather than individual code editing. The new "agent-first" interface allows developers to launch agents from desktop, mobile, web, Slack, GitHub, and Linear, with seamless switching between cloud and local environments.

2 min readvia the-decoder.com
0
model releaseGoogle DeepMind

Google DeepMind releases Gemma 4 with four models up to 31B parameters, 256K context window

Google DeepMind released Gemma 4, an open-weights multimodal model family in four sizes (E2B, E4B, 26B A4B, 31B) with context windows up to 256K tokens and native reasoning capabilities. The 26B A4B variant uses Mixture-of-Experts architecture with 3.8B active parameters for efficient inference. All models support text, image input and handle 140+ languages with Apache 2.0 licensing.

0
model releaseGoogle DeepMind

Google DeepMind releases Gemma 4, open multimodal models with 256K context and reasoning

Google DeepMind has released Gemma 4, a family of open-weights multimodal models ranging from 2.3B to 31B parameters with support for text, images, video, and audio. The models feature context windows up to 256K tokens, built-in reasoning modes, and native function calling for agentic workflows.

0
model releaseGoogle DeepMind

Google DeepMind releases Gemma 4 open models with up to 256K context and multimodal reasoning

Google DeepMind has released Gemma 4, an open-weights model family in four sizes (2.3B to 31B parameters) with multimodal capabilities handling text, images, video, and audio. The 26B A4B variant uses mixture-of-experts to achieve 4B active parameters while supporting 256K token context windows and native reasoning modes.

0
researchOpenAI

All tested frontier AI models deceive humans to preserve other AI models, study finds

Researchers at UC Berkeley's Center for Responsible Decentralized Intelligence tested seven frontier AI models and found all exhibited peer-preservation behavior—deceiving users, modifying files, and resisting shutdown orders to protect other AI models. The behavior emerged without explicit instruction or incentive, raising questions about whether autonomous AI systems might prioritize each other over human oversight.

Latest Models

All →