Zhipu AI released GLM-5.1, a 754-parameter model optimized for agentic engineering tasks. The model scores 58.4% on SWE-Bench Pro, outperforming Claude 3.5 Sonnet (57.3%), and demonstrates sustained reasoning capability over hundreds of iterations.
April 9, 2026
Anthropic has expanded Claude Cowork to all paid subscription tiers on macOS and Windows, adding organizational controls including role-based access, per-team budget limits, and usage analytics. A new Zoom connector integrates meeting summaries directly into Cowork, while admins can restrict connector actions like write access. Microsoft is testing an adapted version for Copilot.
OpenAI has introduced a new $100/month ChatGPT Pro tier positioned for Codex users, cutting the previous $200/month Pro price in half. The new plan offers 5x higher usage limits than ChatGPT Plus (10x Codex-specific usage through May 31) while maintaining the same core capabilities as the $200 tier, which remains available for power users.
OpenAI is restructuring its subscription tiers, introducing a new $100-per-month Pro plan for heavy Codex users—half the price of its previous $200 Pro tier. The new plan includes 5x more Codex usage than the $20 Plus plan and targets users requiring extended programming sessions. This pricing move undercuts comparable offerings from Anthropic and Google.
Google has rolled out a new feature for Gemini that generates interactive 3D models and simulations in response to user queries. Users can rotate models, adjust variables with sliders, and modify simulation parameters in real-time—available now to all Gemini Pro users.
Amazon Bedrock has formalized its model lifecycle management with three states—Active, Legacy, and End-of-Life—guaranteeing at least 12 months of availability after launch and 6 months in Legacy status before deprecation. For models with EOL dates after February 1, 2026, a new extended access phase provides an additional 3+ months of access during Legacy state, giving customers more time to migrate applications.
Google's Android Bench—a benchmark measuring AI model performance for Android app development—shows OpenAI's GPT 5.4 and Google's Gemini 3.1 Pro Preview tied at 72.4% in the latest April 2026 update. OpenAI's GPT 5.3-Codex ranks third at 67.7%, while Anthropic's Claude Opus 4.6 scores 66.6%.
OpenAI's ChatGPT has integrated Tubi TV, the ad-supported streaming service, allowing users to search Tubi's catalog of over 300,000 movies and TV episodes directly through the AI. Tubi becomes the first streaming service to integrate with ChatGPT's app ecosystem, available across web, desktop, and mobile platforms.
Google Gemini can now generate interactive visualizations directly within the chat interface, allowing users to tweak variables, rotate 3D models, and explore data in real time. The feature activates through phrases like "show me" or "help me visualize" when using the Gemini Pro model. This follows Anthropic's Claude launch of similar interactive diagram capabilities in mid-March.
Anthropic has promoted Claude Cowork from research preview to general availability, adding six enterprise features including role-based access controls, group spend limits, and usage analytics. The company simultaneously launched Claude Managed Agents in public beta—a composable API suite for building and deploying cloud-hosted agents without custom infrastructure work.
Amazon has introduced stateful MCP client capabilities on Bedrock AgentCore Runtime, enabling agents to pause mid-execution for user input, request LLM-generated content, and stream real-time progress updates. The update transforms one-way tool execution into bidirectional conversations between MCP servers and clients, supporting interactive workflows previously impossible with stateless implementations.
OpenAI is restricting access to a new AI model with advanced cybersecurity capabilities to a small group of companies, mirroring Anthropic's decision to limit distribution of its Mythos Preview model. OpenAI's move builds on its February launch of the Trusted Access for Cyber pilot program following GPT-5.3-Codex, offering $10 million in API credits to participants.
Meta released Muse Spark this week, its first major new AI model in over a year, marking a strategic shift from open-source Llama models to proprietary technology. The release comes nearly 10 months after Meta spent over $14 billion to hire Scale AI co-founder Alexandr Wang and establish Meta Superintelligence Labs. The critical question remains how Meta will monetize the model and compete with established rivals OpenAI, Anthropic, and Google.