How a Developer Created a Music Generation Skill for Yandex Alice
A developer created a skill for Yandex Alice that generates music on command. You say "create a song about the sea" and wait about a minute. The author criticizes Yandex's outdated skill marketplace and demonstrates that live projects based on generative models are still very much possible.
AI-processed from Habr AI; edited by Hamidun News
A Melbourne-based developer created a skill for Yandex Alice that generates music right during a conversation. The command is simple: say "create a song about the sea," wait a minute—and the smart speaker plays it.
Why the Developer Did It
The author is raising a daughter in a Russian-speaking family in Melbourne and wants Russian to be for her not just an everyday language, but one where interesting and modern things happen. He bought two Yandex Alice units (Max and Pro versions) because there are no proper alternatives for Russian in the smart speaker segment. Amazon Echo, Apple HomePod, and Google Nest barely understand Russian at a schoolboy level, while Yandex handles it right out of the box.
The Problem: Marketplace Stuck in Time
When the author opened Yandex Dialogs (the skill marketplace for Alice), disappointment awaited. It's filled with primitive projects from the pre-ChatGPT era: children's math problems, simple role-playing games, fairy tales. A dead product, as the developer puts it. Looking at this, he thought: come on, surely we can make something alive and useful.
How Music Generation Works Technically
The skill uses modern audio generation models. When a user says "create a song about the sea," the system performs several steps in sequence:
- Processes the voice command and converts it to text
- Sends the description to a music generation model
- Receives the generated audio file
- Plays it through the speaker
The entire cycle takes about a minute. Essentially, this is the first practical example of Audio Diffusion or similar technologies in use in Russia's voice assistant ecosystem. Previously, such things were limited to labs and demonstrations; now they're in the hands of end users.
Why It Was Realized in Two Evenings
The short timeline isn't because it's simple—but because the author was well-prepared. He already had a ready-made infrastructure template and experience with two or three similar side projects. Starting from scratch would have required significantly more time. But the fact that the skill turned out simple enough to implement shows that Yandex APIs are accessible for experimentation, and the barrier to entry isn't catastrophically high.
What This Means
This is neither a revolution nor a replacement for music producers. It's a signal that Russian-speaking developers can experiment with modern generative models within an already-existing platform. Instead of a dead marketplace dominated by fairy tales and role-playing games, there could be space for alive, useful projects that genuinely interest users.
- Oct 28, 2026Keystone removes the model name from code: open-source LLM gateway adds 0.64 ms
- Oct 28, 2026Wired Tested 6 AI Agents in a Day: Dots Leads, Gemini Spark Fails at Lunch
- Oct 28, 2026Anthropic Turns Claude Code Into a Platform: Claude Mods Flip the Agent Harness
- Oct 28, 2026Liquid AI releases LFM2.5-24B-A2B: an agent model with 2B active parameters
Key Questions Answered
Why the Developer Did It?
The author is raising a daughter in a Russian-speaking family in Melbourne and wants Russian to be for her not just an everyday language, but one where interesting and modern things happen. He bought two Yandex Alice units (Max and Pro versions) because there are no proper alternatives for Russian in the smart speaker segment. Amazon Echo, Apple HomePod, and Google Nest barely understand Russian at a schoolboy level, while Yandex handles it right out of the box.
What are the key details on The Problem: Marketplace Stuck in Time?
When the author opened Yandex Dialogs (the skill marketplace for Alice), disappointment awaited. It's filled with primitive projects from the pre-ChatGPT era: children's math problems, simple role-playing games, fairy tales. A dead product, as the developer puts it. Looking at this, he thought: come on, surely we can make something alive and useful.
How Music Generation Works Technically?
The skill uses modern audio generation models. When a user says "create a song about the sea," the system performs several steps in sequence: Processes the voice command and converts it to text Sends the description to a music generation model Receives the generated audio file Plays it through the speaker The entire cycle takes about a minute. Essentially, this is the first practical example of Audio Diffusion or similar technologies in use in Russia's voice assistant ecosystem. Previously, such things were limited to labs and demonstrations; now they're in the hands of end
Build & Ship Production AI Agents Without Boilerplate Code
Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.