> ## Documentation Index
> Fetch the complete documentation index at: https://novita.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# August 22, 2025 Product Updates

## **Model APIs Updates**

### **New Features**

* **New Model Available: MiniMax Speech-2.5**

  Novita now features MiniMax's next-generation Text-to-Speech models: **Speech-2.5-HD** and **Speech-2.5-Rurbo**. Speech 2.5 represents a major breakthrough in voice generation, delivering significant advancements in multi-language expressiveness, voice cloning fidelity, and language coverage.

  Speech-2.5-HD (High-Fidelity):The latest HD model, offering exceptional prosody and remarkable voice cloning accuracy.Specifically designed for high-quality audio applications, providing studio-grade, high-fidelity sound.

  Speech-2.5-Turbo:The latest Turbo model, engineered for extreme responsiveness and ultra-low latency.Ideal for high-concurrency, real-time scenarios and includes support for 40 languages.

  Get started now: MiniMax Speech-2.5-hd-preview Text to Speech, MiniMax Speech-2.5-turbo-preview Text to Speech
* **New Model Available: Qwen-Image-Edit**

  Built upon the powerful 20B Qwen-Image model, Qwen-Image-Edit extends its unique text rendering capabilities into the image editing domain, enabling highly precise in-image text manipulation.

  Key Highlights:

  * **Comprehensive Editing:** Supports a full spectrum of edits, from precise local adjustments (adding, removing, or modifying elements while keeping the rest of the image intact) to high-level creative edits (like IP creation, object rotation, and style transfer).
  * **Precise Text Editing:** Directly add, delete, or modify both Chinese and English text within an image while perfectly preserving the original font, size, and style.
  * **SOTA Performance:** Achieves state-of-the-art (SOTA) results on multiple public benchmarks, establishing it as a powerful foundational model for image editing tasks.

  Get started now:[**Qwen-Image-Edit**](https://novita.ai/docs/api-reference/model-apis-qwen-image-edit)
* **New Model Available: Pixverse V4.5 Text to Video, Pixverse V4.5 Image to Video**

  Pixverse 4.5 is the latest AI video model released by Pixverse AI. It introduces major upgrades in cinematic control, multi-image fusion, and complex motion handling.

  Key Highlights:

  * **20+ Cinematic Camera Controls:** Access over 20 camera controls like panning and zooming, all precisely guided by text prompts for professional-grade results.
  * **Multi-Image Reference & Fusion:** Seamlessly fuse elements from multiple reference images with the new Fusion mode for richer, stylistically consistent videos.
  * **Optimized Complex Motion:** Superior handling of complex actions like high-speed movement and group interactions, delivering smoother and more natural animations.

  Pixverse V4.5 supported text-to-video and image-to-video generation.

***

## **Platform Service Updates**

### **New Features**

* **Budget Management Now Live for Teams**

  Our new Budget Management feature is now available, giving you precise control over your team's API spending. This system is designed to help you efficiently manage member-level spending limits, with key capabilities including:

  * **Member-Level Budget Control:** Assign unique spending caps to individual team members.
  * **Smart Enforcement & Alerts:** Automatically prevent cost overruns and receive timely notifications.
  * **Flexible Budget Types:** Implement one-time or recurring (daily, weekly, or monthly) budgets.
  * **Real-Time Adjustments:** Instantly modify budget allocations as your needs change.

  To get started, visit [Billing > Budgets](https://novita.ai/billing/budgets) in your console. For more details, please see our [help documentation](https://novita.ai/docs/guides/budgets).

  **Updated:** August 20, 2025