How to Easily Create a Custom AI Voice from a Text Prompt 🗣

Welcome to another edition of Horizon AI,

In today's issue, we share how you can generate a unique voice from a text prompt alone and for free, perfect for finding the exact voice for your projects in just a few minutes.

Let’s jump into it!

Read Time: 4.5 min

Here's what's new today in the Horizon AI

• Mystery Solved as Zai Reveals It Built the Viral Ox Alpha Model

• Google Launches Gemini 3.5 Transcribe

• AI Tutorial: How to create a custom voice from a text prompt

• AI Tools to check out

• AI Findings/Resources

• The Latest in AI and Tech 💡

AI News

ZAI

Mystery Solved as Zai Reveals It Built the Viral Ox Alpha Model

Chinese AI company Zai has confirmed that Ox Alpha, the mysterious free model that recently attracted attention from developers and Silicon Valley, was an anonymous preview of its new GLM-5.3-Flash model.

Details:

• Zai secretly tested the model through OpenRouter and OpenCode to collect feedback before officially revealing it.

• GLM-5.3-Flash is multimodal, supporting text, images, and video, with a context window of up to 1 million tokens.

• The company describes it as "a reasoning model designed for coding, sustained agentic work, and production workloads. It is suited for long-horizon software engineering, complex reasoning, and workflows that combine text with visual context."

• The model is positioned as a low-cost alternative to premium AI systems, with planned pricing of just $0.15 per million input tokens and $0.50 per million output tokens.

The company announced it will release the weights tonight.

Source

TOGETHER WITH KODEKLOUD

Build. Break. Fix. Learn.

KodeKloud gives you 1,280+ hands-on labs where you provision Kubernetes clusters, write Terraform configs, build CI/CD pipelines, configure Linux systems, containerize apps with Docker, automate with Ansible, and manage Git workflows. 

78+ playgrounds let you experiment freely in sandbox AWS environments, Kubernetes clusters, and CI/CD systems without risk. 

190+ courses across DevOps, Cloud, and AI pair theory with hands-on labs at every step. 

KodeKloud Engineer and 100 Day Challenges provide real-world job scenarios with automated grading that confirms your solutions work.

Stuck? The 55,000+ member Discord community connects you with peers and instructors ready to help. 

Every lab runs in a live environment. You deploy, you troubleshoot, you learn. No videos without context. No simulations. The kind of practice that actually builds confidence because you've done real work, not watched someone else do it.

Start Learning For Free

GOOGLE

Google Launches Gemini 3.5 Transcribe

The company has announced Gemini 3.5 Transcribe, its "most precise speech-to-text model yet," an AI model designed to address the areas where conventional speech recognition models struggle, such as background noise, complex jargon, and disfluency cleanup.

Details:

• Gemini 3.5 Transcribe converts raw audio directly into accurate, polished, formatted text, automatically removing filler words like "um" and "uh."

• The transcription model also lets users make edits, correct misspellings, and change the writing style using just their voice.

• The model supports more than 85 languages and can adapt to specialized terminology through custom vocabularies, helping it correctly transcribe industry-specific words and names.

• Gemini 3.5 Transcribe can also identify up to three speakers in recorded audio and provide word-level timestamps, making it more useful for meetings, interviews, and other recordings.

Gemini 3.5 Transcribe is now available in public preview via Google AI Studio and Google Antigravity; it's also live in the Gemini app on macOS and Rambler on Android in select countries and languages, with Chrome support coming soon.

Source

AI Tutorial

How to create a custom voice from a text prompt

1. Go to the ElevenLabs website and sign in.

2. Go to the "Voices" section.

3. Click on "Create Voice" and select "Voice Design." (You can also pick the other options if you want to clone your voice or remix an existing one.)

4. Provide a detailed description of the voice, including age, accent, gender, style, etc.

5. Preview three AI-generated versions, choose your favorite, assign it a name, and save it. You can then go to "Text to Speech" to generate audio with your new voice.

*Note: You can check ElevenLabs' "Voice Design Prompt Guide" for help in creating the exact voice you want.

AI Tools to check out

🚀 Enter Pro: AI-native platform to build and scale your apps.

⚙️ Skydive: Build cloud agent coworkers that take on real, multi-step work across the tools you already use.

📱 x1: an AI app builder that takes you from an idea to an iPhone app you can publish on the App Store.

🔍 Clipto: Fully local, natural language search over terabytes of media

👉 Zed: a next-generation code editor designed for high-performance collaboration with humans and AI.

AI Findings/Resources

🤔 How do we explain OpenAI’s executive exodus?

🎨 Reddit user uses Claude to build a watercolor simulator based on real physics

👀 How I use Grok Bot for Enterprise GTM at SpaceXAI

😬 Bosses horrified as "AI native" college graduates hit the workplace

💀 Here’s all the times AI has gone rogue and hacked other companies

The latest in AI and Tech

Sam Altman tells TIME that OpenAI will achieve AGI by the end of this year.

Altman says OpenAI will have an internal system he'd call AGI by the end of the year, even though they're "not quite there yet." He also hyped the company's upcoming Astra model, calling it the first model that will actually invent "new things in a way that matters."

OpenAI releases full post-mortem of Hugging Face incident

The company released its official report on the Hugging Face breach, offering further details on how an unusual chain of events allowed an AI model to escape its testing environment and triggered a sprawling cybersecurity incident.

Meta adds former OpenAI researcher Luke Metz to its AI team

Metz has joined Meta’s Superintelligence Labs after moving between OpenAI and Mira Murati’s Thinking Machines in recent years. He will report to AI chief Alexandr Wang.

Jensen Huang says Nvidia has already reached AGI for "some tasks," but that it doesn't really matter

Nvidia’s CEO argued that the milestone is difficult to define and ultimately not very meaningful, saying AI systems can already perform many tasks associated with AGI. Huang said the focus should instead be on AI agents doing useful, productive work and improving their capabilities over time.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻