Google has introduced Gemini CLI Extensions, enabling third-party integrations with its command-line AI system. This launch, featuring extensions from companies like Figma and Stripe, follows OpenAI's integration of apps in ChatGPT. Unlike the curated access in ChatGPT, Gemini CLI extensions are openly available on GitHub, fostering an open ecosystem.

The first extension is Google's Nanobanana image generator, allowing users to generate images directly from the Gemini CLI terminal. With over one million users, Gemini CLI is heavily utilized by software developers, including Google's own codebase development.

Google aims to transform Gemini CLI into an extensible platform, connecting it to various tools within a developer's toolchain, according to Ryan J. Salva, Google's senior director of product management for developer tools.

Qapita, a Singapore-based equity management platform, has raised $26.5 million in a Series B round led by Charles Schwab. This partnership introduces Schwab Private Issuer Equity Services, powered by Qapita, aimed at U.S. startups for managing cap tables and stock plans. Founded in 2019, Qapita helps private companies track ownership and manage employee equity across Southeast Asia and the U.S. With 2,700 companies using its platform, Qapita is expanding its presence in the U.S. through this collaboration. The funding will also support the launch of a fund admin product across multiple markets.

Google is expanding access to Opal, its AI-powered vibe-coding app, to 15 additional countries, including Canada, India, Japan, South Korea, and Brazil. The app enables users to create mini web apps using text prompts.

Early adopters in the U.S. surprised Google with sophisticated and practical apps, prompting the global expansion.

Opal allows users to describe their desired app, and Google's models generate it. Users can customize the app's visual workflow, edit prompts, and add new steps. Apps can be published and shared for others to test.

Google has improved Opal's debugging program, keeping it no-code. Users can run workflows step-by-step, tweak steps in the console, and see errors in context.

Performance improvements have significantly reduced app creation time. Steps can now run in parallel for complex workflows.

Tesla has introduced more affordable 'Standard' versions of its Model 3 sedan and Model Y SUV, priced at $36,990 and $39,990, respectively. These models offer an estimated 321 miles of range but come with fewer features, including the omission of Autopilot and a more minimalist interior. The release aims to boost Tesla's sales after a decline in 2024, though the pricing is higher than initial expectations, particularly the unrealized $25,000 Tesla.

The new 'Standard' models have differences inside and outside, such as manually adjusted steering wheels and side mirrors, fewer speakers, and no second-row touchscreen. External changes include removing the light bar and glass roof found on pricier Model Y variants. This move is part of Tesla's effort to cut costs, but it also strategically positions Tesla against competitors, including Ford, General Motors, Rivian, Lucid Motors, and Slate Auto, amid changing market dynamics and EV tax credit expiration.

Otter.ai is expanding beyond meeting transcription with a new suite of enterprise tools designed to integrate meeting data into company workflows, creating a centralized knowledge base.

The new product suite includes an API for custom integrations with platforms like Jira and HubSpot, an MCP server for connecting Otter data to external AI models, and an AI agent that can search meeting notes and presentations.

CEO Sam Liang aims to transform Otter.ai into a comprehensive system of record for conversations, facilitating information sharing and driving business value.

Despite privacy concerns and a class-action lawsuit regarding unauthorized recording of conversations, Otter.ai believes increased access to information through AI-powered meeting tools is beneficial for companies.

Mastodon is planning to introduce a new feature called 'Packs,' inspired by Bluesky's Starter Packs, to help new users discover curated collections of accounts to follow.

The feature aims to address the 'cold start' problem of joining new social networks by suggesting relevant accounts to follow.

Mastodon emphasizes user control, allowing users to opt out of appearing in Packs and providing notifications when they are included.

The development is a collaboration with fediverse developers on a Fediverse Enhancement Proposal (FEP), with the initial version expected in the Mastodon 4.6 release.