Releases

SambaNova announced its fifth-generation SN50 RDU and rack-scale SambaRack SN50 system for agentic inference, with shipments planned for the second half of 2026.

Cursor Cloud Agents gained computer-use capabilities for interacting with browsers and graphical applications.

Listen introduced Research Agent to plan studies, synthesize interviews, and answer research questions from customer evidence.

Cursor introduced long-running agents designed to execute extended software-engineering tasks in the cloud.

ElevenLabs launched a dedicated public-sector offering for citizen services and government workflows.

Suno added warp markers, Remove FX, alternates, and time-signature support to Studio.

CoreWeave introduced ARENA, a production-scale lab for testing AI workload performance, economics, and operational readiness before deployment.

Perplexity launched Model Council, allowing users to query multiple frontier models and synthesize their agreements and disagreements.

AIP Document Intelligence became generally available for extracting, structuring, and reasoning over documents.

Moonshot opened public beta access to Kimi Claw, a persistent cloud agent workspace.

Genspark updated its workspace around voice-first task delegation and autonomous work completion.

Modal added richer observability and management controls for running Sandboxes.

Cohere launched Model Vault for dedicated, fully managed and isolated model inference.

Agent Bricks Knowledge Assistant became generally available, powered by instructed retrieval and MLflow evaluation.

Mistral AI expanded Vibe into a fuller agentic development environment for software teams.

Cursor launched reusable Skills for packaging task-specific agent instructions and workflows.

Cursor launched specialized subagents for decomposing and parallelizing complex coding work.

Cursor 2.4 introduced subagents, reusable agent skills, and image generation capabilities.

Krea launched a real-time image editing experience that responds interactively while users paint.

Anthropic released Cowork as a research preview, bringing Claude's agentic capabilities to desktop work across files and applications.

Anthropic introduced HIPAA-ready Claude tools and resources for healthcare providers, payers, health-tech companies, and individuals, alongside expanded life-sciences capabilities.

EliseAI highlighted 2025 launches including its Maintenance App, SentimentAI, AI-Guided Tours, Lease Audits, and Fee Transparency Suite.

Perplexity launched a specialized offering for public-safety organizations.

CoreWeave announced it would extend its cloud platform with NVIDIA’s Rubin architecture for the next generation of AI training and inference.

Surge launched a catalog of ready-to-license frontier datasets and RL environments covering tool use, professional work, instruction following, and advanced science.

Shield AI documented its AI analytics platform for scoring and visualizing flight maneuvers.

xAI launched secure team and enterprise Grok plans with no training on customer data, team administration, SSO and SCIM options.

OpenAI opened app submissions and launched an app directory in ChatGPT.

Harvey delivered major performance, personalization, multimodal, and workflow integration upgrades across Assistant and Vault.

CoreWeave expanded Mission Control with Telemetry Relay entering GA, GPU Straggler Detection in preview, and the Mission Control Agent in preview.

Mistral AI launched Vibe CLI, an agentic command-line coding environment powered by Devstral.

Palantir, NVIDIA, and CenterPoint Energy launched Chain Reaction to coordinate AI infrastructure and critical supply chains.

AWS announced policy controls and automated evaluations for production AI agents.

AWS expanded Amazon Connect with production agentic AI for customer-service interactions.

AWS introduced a frontier agent for investigating and resolving cloud operational incidents.

AWS launched 3nm Trainium3-powered EC2 UltraServers for model training and inference.

AWS launched a service for organizations to create customized frontier models using Nova checkpoints and training infrastructure.

AWS expanded Transform to modernize application, database and user-interface layers of Windows workloads.

Perplexity introduced persistent memory for AI assistants to personalize future interactions.

Perplexity relaunched Shopping as a free personalized conversational shopping experience with native purchasing.

Anthropic expanded Claude for Chrome beyond the waitlist to all Claude Max subscribers.

Genspark launched an all-in-one workspace spanning AI Inbox, Teams, AI Sheets 2.0, Enterprise, and autonomous output creation.

Perplexity launched a native Android version of the Comet AI browser.

Replit launched an AI design mode for rapid interactive mockups and websites.

Google launches Antigravity, an agent-first development environment built around autonomous planning, coding and verification.

ElevenLabs launched visual-generation workflows alongside its audio tools.

Perplexity added stronger visibility, confirmations and user controls for Comet Assistant actions.

Baidu upgraded Qianfan with stronger model services and enterprise agent-development infrastructure.

Baidu introduced a self-evolving machine-learning engineering agent.

Databricks introduced ai_parse_document for multimodal PDF and document understanding integrated with Agent Bricks, AI Search and AI/BI.