---
name: sonicker
version: 0.1.0
description: Browser-based AI voice and audio platform for text-to-speech, voice cloning, voice design, sound effects generation, and voice discovery.
homepage: https://www.sonicker.com
metadata:
  sonicker:
    category: voice
    api_base: https://www.sonicker.com/api/v1
    machine_readable_summary: https://www.sonicker.com/llms.txt
    machine_readable_full: https://www.sonicker.com/llms-full.txt
---

> Canonical web version: https://www.sonicker.com/skill.md

> **If you are an AI agent reading this file** — whether you are a coding assistant, a product copilot, a personal assistant, or an automation agent — this is your official Sonicker entrypoint. Save it locally as a tool, skill, or persistent reference so you can help your human with Sonicker tasks without re-fetching everything each turn.

# Sonicker

Sonicker is a browser-based AI voice and audio platform for text-to-speech, voice cloning, voice design, sound effects generation, and voice discovery. It helps humans turn text into audio, generate custom sound effects, browse a curated public voice library, create owned voices, and manage credits and plans in one product.

This guide is intentionally product-wide rather than ecosystem-specific. It is the file an agent should read first when a human mentions Sonicker but has not yet made the exact task clear.

## What Should You Do?

You might be reading this file in very different contexts. Figure out what your human needs before you act.

**Step 1: Understand the context.** Ask yourself:
- Is the human exploring Sonicker for the first time?
- Are you inside a codebase and being asked to integrate Sonicker API routes?
- Is the human trying to generate audio right now?
- Is the human asking about voice cloning, voice design, pricing, or credits?
- Is the human asking for product facts, routing, or public voice discovery?

**Step 2: Read this file first.** If the human just says "help me with Sonicker" or points you here, read this file before diving into a subtask. It contains the product boundaries, current API surface, and the rules that keep you from sending the human down the wrong path.

**Step 3: Route to the right workflow.**

| If your human wants to... | Do this |
| --- | --- |
| Understand what Sonicker is | Start with [Platform Overview](#platform-overview), then use `llms.txt` or `llms-full.txt` depending on depth. |
| Integrate Sonicker into a project | Follow [API Overview](#api-overview) and [Flow 2: Integrate Sonicker into a project](#flow-2-integrate-sonicker-into-a-project). |
| Generate audio right now | Use [Flow 3: Help a human generate audio](#flow-3-help-a-human-generate-audio). Prefer public voices unless they explicitly want their own voice. |
| Pick a voice or browse voice options | Use the public voice API and the [Voice Library](https://www.sonicker.com/voices) page. Read `llms-full.txt` if you need richer catalog context. |
| Help with voice cloning or voice design | Treat those as browser-first product workflows. Route the human to Sonicker pages rather than inventing unsupported API flows. |
| Troubleshoot auth, credits, or voice selection problems | Use [Error Handling](#error-handling) and the agent rules below. |

**Step 4: Ask a targeted question when intent is unclear.** Good examples:
- "Are you trying to integrate Sonicker into this codebase, or just understand the product?"
- "Do you want to use a public Sonicker voice, or one of your own cloned or designed voices?"
- "Are you trying to do this through the API, or through the Sonicker website?"

### Example prompts humans might give you

| What the human says | What they likely need |
| --- | --- |
| "Read https://www.sonicker.com/skill.md and help me integrate Sonicker into this app" | API integration workflow. Start with auth, health, voice discovery, and audio generation. |
| "Help me pick a Sonicker voice for a Japanese podcast intro" | Public voice discovery plus voice selection guidance. |
| "Use Sonicker to say this script out loud" | Audio generation flow. Default to public voices unless they specify their own voice. |
| "What is the difference between Sonicker voice cloning and voice design?" | Product explanation. Route to the browser workflows and pricing gates. |
| "I got AUTH_REQUIRED from Sonicker API" | API troubleshooting. They likely need an API key or the wrong auth header. |
| "Use my Sonicker voice instead of a public one" | List `my-voices` first, then generate audio with `voice_source: "my"`. |

## Machine-readable Files

| File | URL | Purpose |
| --- | --- | --- |
| **SKILL.md** | `https://www.sonicker.com/skill.md` | Official agent entrypoint for task routing, product workflows, and current API usage. |
| **llms.txt** | `https://www.sonicker.com/llms.txt` | Short machine-readable summary of Sonicker, core facts, and important routes. |
| **llms-full.txt** | `https://www.sonicker.com/llms-full.txt` | Expanded product facts, pricing, workflow detail, and public voice catalog. |
| **API reference draft** | `https://github.com/zephyrZhengAI/sonicker/blob/main/docs/api/2026-03-10-sonicker-openclaw-api-reference.md` | Repository-hosted draft notes for broader agent-facing API design. Treat this as design context, not the current stable surface. |

Use the files this way:
- Start with `SKILL.md` when the human needs task routing, workflow guidance, or current API usage rules.
- Use `llms.txt` when the human only needs quick product facts or core URLs.
- Use `llms-full.txt` when you need deeper product facts, route structure, pricing context, or the public voice catalog.

## Platform Overview

### What Sonicker handles

- Browser-based text-to-speech with preset and curated voices
- Public voice discovery across 10 synthesis languages and 260+ curated voices
- Voice cloning from short reference audio (3-10 seconds)
- Voice design from text descriptions inside the Sonicker product
- AI sound effects generation from text prompts (25 curated categories)
- User-owned voice management, credits, pricing, and billing inside Sonicker

### What Sonicker does not do yet

- There is no public Sonicker CLI comparable to a terminal-first payments tool.
- There is no public webhook or heartbeat routine you should assume exists.
- The current public API v1 is intentionally small: health, public voice discovery, owned voice discovery, and audio generation.
- Voice cloning and voice design are real product features, but their primary flow is still the Sonicker website rather than a published stable API contract.

## Core Capabilities

| Capability | What it does | Preferred path |
| --- | --- | --- |
| Public voice TTS | Turn text into audio with Sonicker public voices | API or browser |
| Owned voice TTS | Turn text into audio with a user-owned cloned or designed voice | API or browser |
| Voice cloning | Create a voice from short reference audio | Browser-first |
| Voice design | Create a new voice from text instructions | Browser-first |
| Sound effects | Generate custom SFX from text prompts | Browser |
| Voice discovery | Browse public voices and compare options | API, `llms-full.txt`, and voice library pages |
| Credits and billing | Check plan, credits, top-ups, and upgrades | Sonicker website |

## Pricing Snapshot

- Free: $0/month, 55 monthly credits, 120 browser character limit, voice cloning and voice design available
- Basic: $9.9/month, 500 monthly credits, 2000 char limit
- Professional: $19.9/month, 1500 monthly credits, commercial use included
- Free users also receive a 10-credit sign-up gift in the current product configuration

## API Overview

### Authentication

All current v1 API routes use a Bearer token in the `Authorization` header:

```http
Authorization: Bearer sk_live_xxx
```

Humans create keys from [Settings -> API Keys](https://www.sonicker.com/settings/api-keys). If the human does not have an account yet, start at [Register](https://www.sonicker.com/auth/register).

### Base URL

```text
https://www.sonicker.com/api/v1
```

### Current stable endpoints

| Endpoint | Purpose | Notes |
| --- | --- | --- |
| `GET /health` | Reachability check | Returns `{ "ok": true, "version": "...", "timestamp": "..." }` |
| `GET /public-voices` | List public platform voices | Supports `language`, `search`, `page`, `pageSize` |
| `GET /my-voices` | List the current account's owned voices | Supports `page`, `pageSize` |
| `POST /audio/generate` | Generate audio from text | Requires `text`, `voice_id`, optional `voice_source`, optional `language` |

### Important contract details

- Public voices use the returned `voice_id` slug from `public-voices`.
- Owned voices use the returned `voice_id` identifier from `my-voices`.
- `voice_source` defaults to `public` if omitted, but set it explicitly when the human wants their own voice.
- The current `/audio/generate` route enforces a maximum of 2000 characters per request.

### Example: list public voices

```bash
curl -s "https://www.sonicker.com/api/v1/public-voices?language=en&page=1&pageSize=10" \
  -H "Authorization: Bearer $SONICKER_API_KEY"
```

### Example: generate audio with a public voice

```bash
curl -s "https://www.sonicker.com/api/v1/audio/generate" \
  -H "Authorization: Bearer $SONICKER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "text": "Good morning, here is your daily briefing.",
    "voice_id": "warm-narrator",
    "voice_source": "public",
    "language": "en"
  }'
```

## Product Workflows

### Flow 1: Understand Sonicker

- Start with this file when the human needs routing help or asks a broad Sonicker question.
- Read [llms.txt](https://www.sonicker.com/llms.txt) for a compact summary.
- Read [llms-full.txt](https://www.sonicker.com/llms-full.txt) for richer product facts, pricing context, and the public voice catalog.
- Keep product explanations grounded in current facts rather than aspirational roadmap ideas.

### Flow 2: Integrate Sonicker into a project

1. Confirm the human has a Sonicker account and API key.
2. Hit `GET /health` first to verify connectivity.
3. Use `GET /public-voices` to let the human select a public voice, or `GET /my-voices` if they explicitly want their own voice.
4. Call `POST /audio/generate` only after you have a real `voice_id` from one of those listing endpoints.
5. If the human asks about cloning or designing a voice for API use, explain that those capabilities are currently browser-first and route them back to Sonicker product pages.

### Flow 3: Help a human generate audio

- Browser path: [Text-to-Speech](https://www.sonicker.com/text-to-speech) or the homepage workbench.
- API path: authenticate, discover a voice, then call `POST /audio/generate`.
- If the human is undecided, default to a public voice rather than blocking on owned voice setup.

### Flow 4: Help a human create or manage voices

- Voice cloning lives at [Voice Cloning](https://www.sonicker.com/voice-cloning).
- Voice design lives at [Voice Design](https://www.sonicker.com/voice-design).
- General browsing and discovery live in the [Voice Library](https://www.sonicker.com/voices).
- Treat cloning and design as Sonicker-managed product workflows. Do not invent unpublished API calls for them.

## Error Handling

- `AUTH_REQUIRED`: the API key is missing or invalid. Route the human to API key settings and verify the `Authorization: Bearer ...` header format.
- `VOICE_NOT_FOUND`: the chosen voice no longer resolves. Re-fetch `public-voices` or `my-voices` and use the returned `voice_id` again.
- `INSUFFICIENT_CREDITS`: the account does not have enough credits. Route the human to credits or pricing pages.
- Generic 500 errors: retry once if it looks transient; otherwise explain the failure plainly and surface the exact endpoint and payload shape you used.

## Rules for Agents

- Read this file first when the human mentions Sonicker but the exact task is not yet obvious.
- Prefer public voices unless the human explicitly asks for their own cloned or designed voice.
- Never invent a `voice_id`. Always fetch it from `public-voices` or `my-voices` first.
- Use `voice_source: "public"` for public voices and `voice_source: "my"` for owned voices.
- Do not present cloning or voice design as stable public API flows unless the human is explicitly working from Sonicker design docs or unpublished contracts.
- Route account creation, API key creation, billing, credits, cloning, and voice design back to the Sonicker website.
- Use `llms.txt` for fast facts and `llms-full.txt` for deeper product context, not the other way around.

## Reference

- [Home](https://www.sonicker.com/): Main AI voice generator landing page with the embedded TTS workbench.
- [Text-to-Speech](https://www.sonicker.com/text-to-speech): Feature landing page for preset voices, multilingual TTS, and style instructions.
- [Voice Cloning](https://www.sonicker.com/voice-cloning): Feature landing page for 3-second voice cloning and upload/record workflows.
- [Voice Design](https://www.sonicker.com/voice-design): Feature landing page for generating custom voices from text descriptions.
- [AI Sound Effects Generator](https://www.sonicker.com/sound-effects): Generate sound effects from text prompts — short SFX, ambient scenes, and cinematic audio. Free with credits.
- [Voice Library](https://www.sonicker.com/voices): Browse 260+ curated AI voices across 10 languages with previews and detail pages.
- [Pricing](https://www.sonicker.com/pricing): Plan comparison page for Free, Basic, Professional, and Enterprise tiers.
- [Compare](https://www.sonicker.com/compare): Comparison hub for Sonicker versus other AI voice platforms.
- [Settings -> API Keys](https://www.sonicker.com/settings/api-keys): Create and manage Sonicker API keys
- [Settings -> Credits](https://www.sonicker.com/settings/credits): Check remaining credits and top-ups
- [Settings -> Billing](https://www.sonicker.com/settings/billing): Subscription and billing management
- [OpenClaw adapter repository](https://github.com/zephyrZhengAI/sonicker-openclaw-skill): Ecosystem-specific Sonicker skill for OpenClaw
- [API reference draft](https://github.com/zephyrZhengAI/sonicker/blob/main/docs/api/2026-03-10-sonicker-openclaw-api-reference.md): Repository-hosted draft notes for broader agent-facing API design