xAI is rolling out Connectors for Grok.

GitHub, Notion, Linear, Google, and Microsoft suites are available, as are custom MCP connectors!

TinyFish makes Search and Fetch APIs free for all developers

TinyFish made Search and Fetch free for all developers and AI agents, with no card required and production-ready limits. The tools span REST, MCP, CLI, SDKs, and key integrations, focusing on fast, structured retrieval and lower token use.

🗞 #sponsored @testingcatalog

Perplexity prepares Digest feature for personalized summaries

Perplexity’s hidden Digest feature appears tied to its context layer and Computer platform, pointing to a recurring daily brief built from connected apps. It may sit behind Max or draw from Computer usage limits.

🗞 #perplexity @testingcatalog

Anthropic is working on Orbit, its upcoming proactive assistant

Anthropic appears close to launching Orbit, an opt-in, time zone-aware assistant for Claude and Claude Code. Build references point to personalized briefings from Gmail, Slack, GitHub, Calendar, Drive, and Figma.

🗞 #claude @testingcatalog

ANTHROPIC 🚨: Claude Cowork will get its own proactive assistant called "Orbit".

> Users will get personalized insights from Gmail, Slack, GitHub, Calendar, Drive, Figma, and other apps, which Claude will generate proactively.

> There are also mentions of "Orbit" apps, which users will be able to "deploy."

> "Your deployed Orbit apps. Pin favorites for quick access."

> OpenAI already has ChatGPT Pulse, while both Google and Perplexity are developing their own proactive assistants, too.

> There is a high chance it will be released as Max-only.

Gemini mobile app redesign leaks show upcoming look ahead of Google I/O

Leaked screenshots show Gemini for iOS adopting Apple’s Liquid Glass style, with translucent panels, a cleaner tool sheet, and reorganized navigation. Google may unveil it at I/O alongside new agents, assistant features, and models.

🗞 #gemini @testingcatalog

Google prepares new upgrades for Gemini Flash model

Ahead of Google I/O 2026, signs point to a Gemini 3.2 Flash launch: LM Arena testing, Vertex AI migration notices, and a brief app listing suggest a near-term rollout with faster responses and stronger reasoning at lower cost.

🗞 #gemini @testingcatalog

AI Studio can now use Nano Banana for Image Generation to tweak images on apps generated via AI Studio Build.

Anthropic announced new ready-to-run Claude agent templates for Finance services.

These tools can be used as plugins for Claude Code, Cowork, or via Managed hosted Agents.

Inworld AI launches Realtime TTS-2 model for live conversations

Inworld AI launched Realtime TTS-2, a voice model for live conversation with sub-200ms latency, conversational context, natural-language voice direction, inline vocal controls, and voice consistency across 100+ languages.

🗞 #sponsored @testingcatalog

OPENAI 🚨: GPT-5.5 Instant is rolling out to all users on ChatGPT! "gpt-5.5-chat-latest" is coming to APIs as well.

> Much more concise. Better memory. More personalized.

Instant testing time 👀

Perplexity released Perplexity Computer for Professional Finance with bootstrapped workflows and new data providers.

Additionally, all responses are traceable back to the source!

Perplexity got a new tab for artifacts! It appears to be a list of all previously generated artifacts across Perplexity and Perplexity Computer.

Pinning is possible too 👀

Google released Pomelli Catalog, a new feature on Pomelli marketing agent experiment.

Pomelli will generate a set of products based on your business DNA, so you can reuse them across your marketing campaigns.

For TestingCatalog, it generated a Weekly Newsletter Subscription product and several others.

I will package a new campaign soon 👀

Google released Multi-Token Prediction (MTP) drafters for the Gemma 4 family. It comes with a 3x speed boost without losing performance.

Looking forward to testing a quantized Gemma 4 with MTP drafters on a Mac Mini!