Sign in Subscribe
2 min read

Anthropic Launches Opus 4.7 as Microsoft Takes Lead on OpenAI’s Stargate Project

Model Releases and Updates

Anthropic Releases Opus 4.7 Across Platforms: Anthropic has launched Opus 4.7, an updated version of its high-end model series, which has been deployed on both Google Vertex and the Claude Web interface. This release indicates ongoing iterations to the Opus line, providing users with enhanced capabilities for both enterprise and consumer applications.

Gemini 3.1 Pro Leads METR Timeline Benchmarks: Google’s Gemini 3.1 Pro has claimed the top position on the METR Timeline benchmark, achieving an 80% success rate on tasks lasting 1.5 hours. The model's performance in handling long-duration tasks suggests significant advancements in reliability and complex reasoning.

1-bit Bonsai 1.7B Enables Efficient Local Inference: Bonsai 1.7B is a new 1-bit AI model with a 290MB footprint designed to run locally in web browsers using WebGPU. This development advances the accessibility of AI by allowing efficient, serverless execution of models on local hardware.

Developer Tools and Infrastructure

Mistral AI Launches Public Preview of Connectors API: Mistral AI has introduced its Connectors API in public preview, utilizing the Model Context Protocol (MCP) to allow developers to build integrations once and deploy them across various platforms. The tool is designed to enhance efficiency and flexibility in how AI models interact with external data sources.

Mistral Small 4 Shows High Performance in Agentic Workflows: Users report that running Mistral Small 4 through the Hermes agent harness and Open WebUI significantly outperforms the native Le Chat interface. This setup enables superior multi-step tool use and greater control over system prompts, highlighting the model's potential for complex agent loops.

Microsoft Takes Over OpenAI’s Stargate Norway Project: OpenAI has withdrawn from its data center agreement in Norway, with Microsoft stepping in to lead the "Stargate" project. This transition marks a strategic shift in OpenAI's infrastructure plans and underscores Microsoft's increasing role in managing large-scale AI compute facilities.