Capture your thoughts at the speed of sound.
Vocalog is an Obsidian plugin that brings one-click live voice recording, lightning-fast Whisper speech-to-text (STT), and LLM-powered intelligent summarization directly into your daily notes and PKM workflow.
Most voice note solutions require you to record in an external app, export the audio, find a transcription service, paste the transcript into Obsidian, and manually organize it.
Vocalog automates the entire pipeline inside Obsidian:
- Speak: Click the ribbon mic icon to start live recording.
- Transcribe: Ultra-fast transcription via OpenAI Whisper, Groq, or local models.
- Synthesize: An LLM (DeepSeek, GPT-4o, Claude, Ollama) distills your speech into structured daily notes (actions, insights, summaries).
- Preserve: Raw verbatim transcripts are automatically kept in collapsible blocks so you never lose context.
- ๐๏ธ In-App Live Microphone Recording: Record voice memos directly inside Obsidian without external tools.
- ๐ด Active Visual Pulse Feedback: Ribbon button pulses with glowing red animation while recording is active.
- ๐ค Custom AI Mic Studio Icon: Specially designed icon representing voice intelligence.
- โก Whisper Speech-to-Text Support: Compatible with any OpenAI-compatible audio transcription API (Groq Whisper, OpenAI Whisper, LocalAI, self-hosted).
- ๐ง Smart LLM Summarization: Turns noisy speech into actionable Markdown journals, extracting TODOs, insights, and key points.
- ๐ Raw Transcript Preservation: Raw transcripts are retained in collapsible
<details><summary>๐ Raw Transcripts</summary></details>blocks. - ๐
Automated Daily Journaling: Appends entries cleanly to your existing daily notes and refreshes frontmatter
updated:timestamps. - ๐๏ธ Interactive Calendar & Date Range Batching: Batch process historical recordings for any selected days or date ranges.
- ๐ 100% Private & Local-First: Zero telemetry, zero tracking, no intermediate servers. All API keys and recordings remain strictly in your local vault.
For the fastest and most cost-effective experience:
| Service | Provider | Recommended Model | Benefits |
|---|---|---|---|
| Speech-to-Text (STT) | Groq Cloud | whisper-large-v3 / whisper-large-v3-turbo |
Near-instant transcription speed, generous free tier |
| Summarization (LLM) | DeepSeek or OpenAI | deepseek-chat / gpt-4o-mini |
High-quality reasoning, highly affordable, concise Markdown |
- Download the latest
main.js,manifest.json, andstyles.cssfrom the Releases page. - Create a folder in your vault:
<vault>/.obsidian/plugins/vocalog/ - Copy the downloaded files into that folder.
- Open Obsidian โ Settings โ Community plugins โ Enable Vocalog.
Navigate to Settings โ Vocalog:
- Audio Folder: Folder where recordings are stored (e.g.
attachmentsorInbox/Voice). Defaults toInbox/Voice. - Output Folder: Folder where daily notes are saved (leave empty for vault root).
- Daily Note Format: Date pattern matching your daily notes (default:
YYYY-MM-DD).
- STT API URL:
- Groq:
https://api.groq.com/openai/v1/audio/transcriptions - OpenAI:
https://api.openai.com/v1/audio/transcriptions
- Groq:
- STT API Key: Your provider API key.
- STT Model:
whisper-large-v3(Groq) orwhisper-1(OpenAI).
- LLM API URL:
- DeepSeek:
https://api.deepseek.com/v1/chat/completions - OpenAI:
https://api.openai.com/v1/chat/completions - Groq:
https://api.groq.com/openai/v1/chat/completions - Local (Ollama):
http://localhost:11434/v1/chat/completions
- DeepSeek:
- LLM API Key: Your LLM API key.
- LLM Model:
deepseek-chat,gpt-4o,llama-3.3-70b-versatile, etc. - System Prompt: Customize how the AI formats your daily notes.
- Click the Vocalog Studio Mic icon in the left ribbon to start recording.
- The icon pulses red while you speak.
- Click the icon again to stop: Vocalog automatically transcribes, generates AI notes, and writes to your daily journal!
Vocalog: Start / stop live audio recordingVocalog: Generate audio notes(processes today's audio files in the folder)Vocalog: Generate audio notes by date range(opens interactive calendar modal)Vocalog: Generate from selected audio files(processes files selected in File Explorer)
- Right-click any
.mp3,.m4a,.wav,.webm, or.oggfile in Obsidian โ select Generate audio note.
- Direct Connections: Network requests are made directly from your Obsidian client to your configured API endpoints.
- No Third-Party Middleman: There are no proxy servers, tracking scripts, or analytics.
- Credential Storage: API keys are securely saved only in your local
.obsidian/plugins/vocalog/data.jsonfile.
Contributions, feature requests, and bug reports are warmly welcome!
- Open an Issue to report a bug or request a feature.
- Submit a Pull Request following standard Git Flow (
feature/...branches intodevelop).
Distributed under the MIT License.