--- title: Google Gemini AI video module update, TypeSafe, and more app updates slug: google-gemini-ai-video-module-update-typesafe-and-more-app-updates description: Explore Google Gemini AI video model changes, a new TypeSafe app, WhatsApp Business Cloud pricing, and Claude Sonnet 5.5 support. docTags: createdAt: 2026-10-02T08:50:09.682Z --- Here's a quick look at our app updates this week. ## Google Gemini AI: Generate a video module update required by September 30, 2026 Google deprecated the gemini-omni-flash-preview model, so the **Generate a video** module now uses the new gemini-omni-1.1-flash model, adding control over video output resolution and generation task type. Scenarios using the previous model stopped working after September 30, 2026. ### What's changing? - The **Generate a video** module now uses the new gemini-omni-1.1-flash model, replacing the deprecated gemini-omni-flash-preview model. - New features include: - **Output resolution**: Select 360p, 720p, 1080p, or 4K for your video output. - **Generation task**: Specify edit or extend for more precise control. - Existing scenarios using the deprecated model continued to run until September 30, 2026. - You can no longer select the deprecated gemini-omni-flash-preview model in new modules. ### What do you need to do? To keep your scenarios running, update your existing **Generate a video** modules: 1. Open any scenario that uses the Google Gemini AI **Generate a video** module. 2. The module automatically defaults to the new gemini-omni-1.1-flash model. Save the escenario to apply the update. ### Resources - [Google Gemini AI documentation β€” Help Center](https://apps.make.com/NrUNDw84SHdHFeIeX-yKe) - [Google Gemini API changelog](https://ai.google.dev/gemini-api/docs/changelog) - [Google Gemini Omni guide](https://ai.google.dev/gemini-api/docs/omni) *** ## WhatsApp Business Cloud: App updates for new Meta pricing We've updated the WhatsApp Business Cloud app to align with Meta's new pricing model for service messages, which takes effect on October 1, 2026. These changes help you monitor your usage and prevent messages from failing to send. **Updated modules** - **Watch Events:** A new `Type` field has been added to the `Pricing` collection in the webhook's output. This field indicates whether a message was part of your free monthly tier (`free_customer_service`) or paid (`regular`), allowing you to monitor when you have used your free message allowance. - **Send a Message:** The information banner in the module has been updated. It now clarifies that after your 1,000 free monthly service messages are used, subsequent messages will not be delivered unless a valid payment method is configured in your Meta Billing Hub. **Resources** - [WhatsApp Business Cloud documentation β€” Help Center](https://apps.make.com/whatsapp-business-cloud) *** ## TypeSafe is now available on Make TypeSafe is a decision-making API that evaluates content and provides structured answers with confidence scores. You can now connect it to your Make scenarios to automatically classify, route, score, and gate your workflows. ### Available modules The TypeSafe app includes the following modules: - **Evaluate a state:** Evaluates state against typed questions and gets back structured answers. - **Make an API Call:** Sends a custom API call to TypeSafe. You can use this to call endpoints that aren’t covered by existing modules. ### Who it's for This integration is useful for support and operations teams who want to automate the routing of incoming requests, or for product teams looking to systematically score customer feedback. ### Resources - [TypeSafe documentation β€” Help Center](https://apps.make.com/typesafe) - [TypeSafe API documentation](https://docs.typesafe.ai/api) - [Community post - πŸ”₯ Feature Spotlight: New TypeSafe App and Jev AI Model in Make](https://community.make.com/t/feature-spotlight-new-typesafe-app-and-jev-ai-model-in-make/116185) *** ## Improved App Search: Find the TypeSafe app by searching for 'Jev' Searching for the AI model 'Jev' in the app gallery or Scenario Builder previously returned no results, making the TypeSafe app hard to find. You can now find the TypeSafe app by searching for 'Jev' in addition to 'TypeSafe'. This helps AI developers and automators quickly find the app using the model name they already know. *** ## Anthropic (Claude): App adds support for Claude Sonnet 5.5 model We've updated the Anthropic (Claude) app to include support for the latest model, **Claude Sonnet 5.5**, available for users on paid plans. ### Updated modules - **Simple Text Prompt (via Amazon Bedrock):** You can now select **Claude Sonnet 5.5** from the **Model** dropdown to leverage its advanced capabilities in your scenarios. - **Create a Prompt:** This module has been updated with several enhancements: - **New model available:** You can now select **Claude Sonnet 5.5** from the **Model** dropdown. - **New diagnostics output:** The module's output now includes a `diagnostics` parameter, offering more detailed information about the model's execution. - **Tool Choice guidance:** When using the Sonnet 5.5 model, tool choice is limited to the **Auto** setting. To prevent errors, the **Any** and **Tool** options now include a description clarifying they are not supported by Sonnet 5.5. ### Resources - [Anthropic (Claude) documentation β€” Help Center](https://apps.make.com/GSbB58XXBpYArJF15YrxB) *** ## Detrack: Send job notifications to multiple emails The **Notify Email** field in the **Detrack > Create a Job** module previously accepted only a single email address. You can now enter multiple email addresses as a text string to send job notifications to several recipients at once. *** ## TimeCamp: 'Make an API Call' module issue resolved We've resolved the following issue in the TimeCamp app: - **TimeCamp:** An issue with the **Make an API Call** module that prevented new tasks from being created has been resolved. The module was communicating with an outdated API endpoint, which caused it to fail silently. It now connects to the correct endpoint, and creating tasks works as expected.