Releases

Anthropic acquired Stainless, an SDK and MCP server tooling company that had powered generation of Anthropic's official SDKs.

Scale marked its tenth anniversary, reporting more than 700 customers served, 15 billion human decisions applied to model development, work across 150 languages and locales, and over $1 billion paid to contributors.

Perplexity added data analysis, coding and visualization workflows for business teams to Computer.

CoreWeave launched secure isolated sandboxes for reinforcement learning, agent tool use, and model evaluation at scale.

Perplexity published details of the security architecture and safeguards built into Computer.

Artificial Analysis data placed CoreWeave first for Kimi K2.6 inference throughput and price-performance among evaluated providers.

Baidu released a more compact flagship model with stronger agents, reasoning and creative abilities.

Scale completed SWE Atlas by adding Test Writing and Refactoring alongside Codebase QnA, producing a 284-task software-engineering evaluation suite.

Perplexity broadened Personal Computer availability to all users on Mac.

The U.S. defense department increased the potential ceiling of its Scale production agreement from $100 million to $500 million following rapid adoption across military components.

Perplexity added structured financial search and retrieval to the Agent API.

SpaceXAI added deep app connectors to Grok, followed by persistent Skills for documents, decks, spreadsheets and workflows.

SpaceXAI agreed to provide Anthropic access to Colossus 1, with both companies also expressing interest in multi-gigawatt orbital AI compute.

The former xAI operation ceased functioning as a separate company and was folded into SpaceX as its AI division, using the SpaceXAI identity.

Anthropic agreed to use all compute capacity at SpaceX's Colossus 1 data center, adding more than 300 megawatts and over 220,000 NVIDIA GPUs.

Harvey launched more than 500 ready-to-use legal agents across major practice areas.

ElevenLabs reported surpassing $500 million in annual recurring revenue during the first four months of 2026.

IBM announced an agentic control plane for deploying and governing agents from multiple sources.

Perplexity launched specialized professional-finance workflows in Computer.

Perplexity added premium clinical and health-information sources to its health experience.

Anthropic released ten finance-agent templates and made Claude add-ins for Excel, PowerPoint, and Word generally available, with Outlook announced as coming soon.

Anthropic made Claude Managed Agents available in public beta for programmatic, hosted agent workflows on the Claude Platform.

Perplexity launched Computer inside Microsoft Teams.

Anthropic, Blackstone, Hellman & Friedman, and Goldman Sachs announced a new AI services company focused on deploying Claude at mid-sized enterprises.

Scale signed a memorandum of understanding with the U.S. Department of Energy to explore AI-ready scientific data infrastructure, evaluations, and model applications for the Genesis Mission.

Anthropic disclosed that annualized run-rate revenue had crossed $47 billion earlier in May 2026.

Cursor described its continuously improving agent harness and the feedback loops used to develop coding agents.

CoreWeave added guided self-service and SUNK Anywhere so research teams could deploy production-grade Slurm-on-Kubernetes training on CoreWeave and other clouds.

CoreWeave launched private cross-cloud networking alongside SUNK Anywhere and LOTA Cross-Cloud to simplify distributed AI workloads.

LOTA Cross-Cloud extended CoreWeave’s local object-transport acceleration to data held outside CoreWeave regions.

Perplexity expanded Computer with connected workplace applications and enterprise workflows.

SpaceXAI launched voice cloning from short recordings for text-to-speech and voice-agent APIs.

ElevenLabs launched ElevenMusic as a dedicated music-creation application.

IBM published the 8B base and instruct Granite 4.1 models.

IBM published the 30B base and instruct Granite 4.1 models.

IBM published the 3B base and instruct Granite 4.1 models.

IBM released Granite 4.1 language, vision, speech, embedding and Guardian models under Apache 2.0.

AWS expanded Quick with a desktop application, Free and Plus plans, visual-asset generation and new integrations.

AWS announced a managed OpenAI-powered agent runtime integrated with Bedrock and AWS controls.

AWS and OpenAI expanded their partnership to offer OpenAI frontier models through Amazon Bedrock in limited preview.

NVIDIA released an open multimodal model unifying vision, audio and language for efficient agents.

Palantir renamed AIP Agent Studio to AIP Chatbot Studio and continued expanding governed agent-building capabilities.

Mistral AI opened Workflows in public preview, enabling visual composition of enterprise AI agents and automations.

Meta agreed to deploy tens of millions of AWS Graviton cores for CPU-intensive agentic AI workloads.

OpenAI and Microsoft announce the next phase of their partnership with amended commercial terms.

OpenAI and Microsoft announced the next phase of their partnership with amended commercial terms.

OpenAI discontinued the standalone Sora product as its video strategy shifted to newer experiences and models.

The efficient V4 Flash preview became available through DeepSeek's API.

DeepSeek introduced the V4 family through Pro and Flash preview API models.

The V4 Pro preview became available through OpenAI- and Anthropic-compatible interfaces.

Cohere and Aleph Alpha announced plans to combine into a transatlantic sovereign-AI company operating as Cohere; completion remained subject to approvals.

SpaceXAI released its first named Think Fast speech-to-speech voice-agent model.

OpenAI released GPT-5.5 and GPT-5.5 Pro as upgraded frontier models for complex knowledge work and reasoning.

Perplexity published its approach to building and evaluating accuracy in frontier AI systems.

Google Cloud announces that future Vertex AI services and roadmap evolution will be delivered through Gemini Enterprise Agent Platform.

Google Cloud launches Gemini Enterprise Agent Platform as the evolution of Vertex AI for building, scaling and governing AI agents.

Anthropic committed more than $100 billion to AWS over ten years for up to five gigawatts of Trainium capacity; Amazon committed $5 billion immediately and up to $20 billion more.

OpenAI released ChatGPT Images 2.0 with stronger world knowledge, instruction following, detail, and an images-with-thinking mode.

Kimi added agentic presentation and slide creation capabilities.

Kimi expanded its workspace features for producing and editing work artifacts.

Moonshot documented Kimi Researcher as a deep-research workflow within its agent suite.

Kimi expanded its multi-agent swarm system to coordinate parallel specialist agents.

Moonshot released Kimi K2.6 with improved agent capabilities and published model weights.

Scale acquired ICG Solutions and its LUX real-time streaming-data analytics platform to expand its national-security AI stack; ICG initially remained a wholly owned subsidiary.

Anthropic disclosed that annualized run-rate revenue had surpassed $30 billion, up from about $9 billion at the end of 2025.

Amazon made an additional $5 billion investment in Anthropic and said it could invest up to a further $20 billion.

Anthropic committed more than $100 billion over ten years to AWS technologies in an agreement securing up to five gigawatts of compute capacity for Claude.

SpaceXAI announced its speech-to-text and text-to-speech APIs with multilingual support and developer pricing.

Anthropic Labs launched Claude Design for creating polished designs, prototypes, slides, and other visual work.

Perplexity launched Personal Computer on Mac for local-file, application and browser workflows.

Anthropic released Claude Opus 4.7 with improved software engineering, vision, complex multi-step work, and new cybersecurity safeguards.

Baidu released an open-weight 8B DiT text-to-image model.

SpaceXAI launched batch and streaming speech transcription in 25 languages.

Google launches Gemini 3.1 Flash Audio for real-time and batch audio understanding and generation.

Midjourney released V8.1 with faster rendering, stronger prompt adherence, and native 2K HD images.

Cursor released Composer 2, a specialized coding model trained for long-horizon agentic software engineering.

Jane Street agreed to purchase $6 billion of CoreWeave AI cloud services and invest $1 billion in CoreWeave equity.

Cursor 3.1 upgraded voice input for directing coding agents.

Cursor 3.1 introduced a tiled multi-agent layout and upgraded voice input.

ElevenLabs introduced local and on-premise deployment options for enterprise voice AI.

CoreWeave announced a multi-year agreement to support development and deployment of Anthropic’s Claude model family.

Perplexity added personal financial-analysis and CFO-style workflows to Computer.

Meta and CoreWeave announced an expanded $21 billion agreement under which Meta would use CoreWeave Cloud to scale AI inference workloads.

Meta Superintelligence Labs introduced Muse Spark, a multimodal model designed for efficient broad intelligence.

Modal added NVIDIA RTX PRO 6000 Blackwell GPU instances.

Modal opened a beta filesystem API for programmatic sandbox file operations.

Anthropic released Claude Mythos Preview as a gated frontier model for vetted Project Glasswing partners, focusing its advanced cybersecurity capabilities on defensive use.

Anthropic launched Project Glasswing with major technology and infrastructure partners to use frontier AI for defensive cybersecurity and critical-software protection.

Cursor launched a dedicated Agents Window for orchestrating multiple coding agents.

Cursor launched Design Mode for visually iterating on user interfaces with agents.

Cursor 3.0 introduced the Agents Window, parallel work across local, worktree, cloud, and SSH environments, and Design Mode.

Perplexity launched tax research and preparation workflows in Computer.

Microsoft makes MAI-Voice-1 available to developers in Microsoft Foundry with custom-voice support and safety controls.

Microsoft AI launches MAI-Image-2 for faster, higher-quality image generation in Copilot and Microsoft Foundry.

Microsoft AI launches MAI-Transcribe-1 for high-speed multilingual speech-to-text through Microsoft Foundry.

Google releases Gemma 4 open models and makes them available through Google Cloud's AI development stack.

Baidu upgraded its self-evolving agent and reported state-of-the-art MLE-Bench performance.

Baidu expanded Miaoda with enterprise and mobile editions and standalone application generation.

Baidu upgraded GenFlow's multi-agent productivity capabilities.

Perplexity launched Computer inside Slack for completing work from shared channel context.