- Horizon AI
- Posts
- How to Use Gemini to Generate Original Music Tracks 🎵
How to Use Gemini to Generate Original Music Tracks 🎵
Anthropic releases Sonnet 4.6 🔥
Gina Acosta
February 18, 2026
In partnership with
Welcome to another edition of Horizon AI,
In today's issue, we look at Gemini's new music creation capability that lets you create short music tracks directly inside the app. You can generate instrumentals or songs with lyrics using not only text prompts, but even media.
Let’s jump into it!
Read Time: 4.5 min
Here's what's new today in the Horizon AI
Anthropic Launches Claude Sonnet 4.6 with Improved Coding, Computer Use, and Office Tasks
Google Brings Music-Generation Capabilities to Gemini App
AI Tutorial: How to Create Music Tracks with Gemini
AI Tools to check out
AI Findings/Resources
The Latest in AI and Tech 💡
AI News
ANTHROPIC
Anthropic Launches Claude Sonnet 4.6 with Improved Coding, Computer Use, and Office Tasks
Anthropic has rolled out Sonnet 4.6, the latest update to its midsized model line, focusing on stronger coding ability, instruction following, and computer use.
Details:
Sonnet 4.6 now becomes the default model for free and pro users.
The beta version introduces a 1 million token context window, 2x the size of the largest window previously available for Sonnet and, according to Anthropic, "enough to hold entire codebases, lengthy contracts, or dozens of research papers in a single request.
On benchmarks, it shows significant improvements across the board, with standout results including 79.6% on SWE-Bench Verified for coding skills (just slightly below Opus 4.6's 80.8%), an impressive 60.4% on ARC-AGI-2 (meant to measure skills specific to human intelligence), and 72.5% on OSWorld for computer use.
Anthropic also notes that it is better at instruction following and is now capable of handling office tasks that previously required an Opus model.
The release arrives two weeks after Opus 4.6, with an updated Haiku model expected to follow in the coming weeks.
TOGETHER WITH AVALARA
Want clarity on sales tax automation? Take the Avalara AvaTax product tour.
Sales tax is complex, but evaluating software shouldn't be. Avalara’s self-guided tour of AvaTax shows how to calculate tax in real time, track changing nexus thresholds, and apply the correct rates across states and jurisdictions.
Used by more than 43,000 companies, AvaTax connects to platforms like NetSuite, QuickBooks, and Shopify, helping you automate compliance across more than 1,400 systems. You’ll see how to review flagged transactions by jurisdiction, create audit-ready reports, and improve accuracy without needing to meet with sales.
If you’re ready to explore automation but short on time, this tour is designed for you.
Google Brings Music-Generation Capabilities to Gemini App
Google is bringing AI music creation directly into the Gemini app, introducing a beta feature powered by Google DeepMind's Lyria 3 model.
Details:
The update lets anyone generate 30-second tracks from text prompts or media. Users can also customize elements like style, vocals, and tempo.
The company said that Lyria 3 is a significant improvement over previous versions, creating more realistic and complex music tracks.
Google is also making the model available to YouTube creators through the Dream Track feature, a tool that helps YouTube creators make AI-generated tracks.
All songs created with the Lyria 3 model will have a SynthID watermark to identify AI-generated content. Users can also upload tracks to Gemini to check if they are AI-generated.
Music generation is available today in the Gemini app for all users 18+ (subscribers will enjoy higher limits) in English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese, with plans to expand to more languages.
AI Tutorial
How to Create Music Tracks with Gemini
Go to the Gemini website, click on "Tools" and select "Create Music"
Describe the song you want to create
For example: "Create a classic 80s rock song about a road trip across the desert with friends. Include electric guitar solos, driving drums, and lyrics about freedom and adventure."
While Google said that you can't mimic an artist outright, you can still add an artist's name to your prompt to have Gemini create a track in a similar style or mood.
Example: "Generate a gritty country song in the style of Johnny Cash with raw storytelling, a steady train-beat rhythm, and his signature deep voice."
You can also upload a photo or a video, and Gemini will create a song to match the mood of the media file.
Gemini will generate a 30-second track with lyrics or instrumental audio and leverage Nano Banana to create custom cover art. You can download and share the result.
Note: The feature is only available on desktop for now. It will roll out for the mobile app "over the next several days"
AI Tools to check out
🌌 SciSpace: End-to-end platform for discovering, analyzing, and writing scientific literature
🎶 Mureka V8: AI music generator that creates unique, customizable songs, lyrics, and tracks for any project, now delivering richer, more natural, and expressive music.
🚀 TheTabber: It helps creators upload or pull content from another account, then repurpose and publish it everywhere.
🍋 HeyLemon: The first Al agent that turns your voice instructions into finished tasks.
🗣️ PolyAI: Conversational AI platform delivering both unlimited scale and complete control, proven at enterprise level.
TOGETHER WITH WISPR FLOW
Ship the message as fast as you think
Founders spend too much time drafting the same kinds of messages. Wispr Flow turns spoken thinking into final-draft writing so you can record investor updates, product briefs, and run-of-the-mill status notes by voice. Use saved snippets for recurring intros, insert calendar links by voice, and keep comms consistent across the team. It preserves your tone, fixes punctuation, and formats lists so you send confident messages fast. Works on Mac, Windows, and iPhone. Try Wispr Flow for founders.
AI Findings/Resources
📌 Every lesson I learned from 210+ hours using OpenClaw this past month
👀 I hacked ChatGPT and Google's AI – and it only took 20 minutes
🤖 China's humanoid robots take centre stage for Lunar New Year showtime
⚠️ Google says attackers used 100,000+ prompts to try to clone AI chatbot Gemini
The latest in AI and Tech
The new feature converts AI-generated code into fully editable designs inside Figma, creating a workflow bridge between AI coding agents and design teams
Elon Musk says xAI has opened a public beta for Grok 4.20, promising major speed and intelligence gains over the prior version.
The release positions the chatbot as a higher-level assistant capable of reviewing medical files and tackling engineering questions, while Musk notes daily bug fixes during the beta period.
Elon Musk’s automaker is rolling out xAI’s Grok chatbot to vehicles in the U.K. and eight additional European markets, integrating the AI assistant into its infotainment systems.
WordPress.com has introduced an AI Assistant designed to help users edit and manage their websites using natural language commands.
The assistant works directly inside your site, allowing you to adjust layout, styles, colors, fonts, and content simply by describing what you want.
China is ramping up support for brain-computer interface research to build domestic rivals to Neuralink, with startups like NeuroXess moving rapidly toward human trials.
Backed by looser regulation, state funding, and a national roadmap designating BCIs as a strategic sector, NeuroXess reports early success enabling a paralyzed patient to control a cursor days after implantation.
That’s a wrap!
We'd love to hear your thoughts on today's email!Your feedback helps us improve our content |
Not subscribed yet? Sign up here and send it to a colleague or friend!
See you in our next edition!
Gina 👩🏻💻