WhisperDirect vs Resonance AI Notes

WhisperDirect is an audio transcription and summarization app built around the high accuracy of the Whisper API. In addition to the Whisper API using your own API key, it includes on-device features such as Apple Speech, Speaker Diarization (NVIDIA Nemotron 3), and OCR. For summaries and meeting minutes, choose from OpenAI / Gemini / OpenAI-compatible LLMs, using APIs only as much as you need. One-time purchase — no subscription. Pricing & Experience ・7-day free trial included ・On-device features (Apple Speech / Speaker Diarization / OCR) are free, no API key required ・Whisper API and LLM (Summary / Minutes) are billed by each provider on a usage basis 　(The app does not charge for API usage) Cost Estimates ・Whisper API: ~$0.006/min (1 hour ≈ $0.36) ・Segment re-transcription (gpt-transcribe): ~$0.0045/min ・On-device features incur no API fees ・Actual costs vary with audio length and settings ・OpenAI API pricing: https://openai.com/api/pricing ・Gemini / custom providers follow each service's pricing ・Locally-run custom providers (e.g. Ollama) incur no API fees Transcription Engines ・Whisper API (OpenAI): high accuracy with your own API key ・Apple Speech (On-device): no API key, iOS 26+ ・Switchable from settings Models for Summary & Minutes ・Choose from OpenAI / Gemini / Custom 　- OpenAI: GPT-6 Luna / GPT-5 nano / GPT-5 mini / GPT-4.1 nano / GPT-4.1 mini 　- Gemini: requires Google AI Studio API key (gemini-*-pro / flash / gemma) 　- Custom: OpenAI-compatible APIs (Ollama / LM Studio / self-hosted) ・Processing costs only a few cents even for several-thousand-character text Key Features ・Record with the mic button and transcribe on the spot ・Live Preview: real-time transcription while recording (on-device) ・Recording display mode switch (Waveform / Simple) ・Import audio files → transcribe (Share Sheet) ・Video import (audio extracted and compressed) ・Transcript auto-highlights with playback position ・Insert timeline markers at any interval (5-sec steps) ・Generate summaries from transcripts ・OCR from images (batch supported, no API fees) ・Generate meeting minutes (prompts editable) ・Re-transcribe misrecognized segments in the editor 　(Uses OpenAI gpt-transcribe; billed by usage) ・Re-transcribe specific items with Whisper API, even while using Apple Speech ・Speaker diarization (NVIDIA Nemotron 3, on-device) ・Export: Audio / Text / Summary / Minutes / VTT / SRT ・Auto-post to Slack ・Webhook integration (Discord, Slack, n8n, etc.) ・Google Drive auto-backup ・Estimated cost display ・Fine-grained settings (LLM model, timeline interval, prompts) Supported Formats Audio: mp3, m4a, aac, wav, flac, ogg, opus, wma, amr, mpga, webm, aiff, caf Video: mp4, mov, m4v, webm, mkv, avi, mpeg, mpg Notes ・Whisper API requires an OpenAI API key ・On-device features work without an API key ・Prices, specs, and models may change per provider

Never lose track of what was said or decided. Resonance records your meetings, transcribes every word, and uses AI to produce structured notes with summaries, key decisions, action items, and open questions: automatically. Whether you are running a one-on-one, a team standup, or a client call, Resonance turns spoken conversation into organized, searchable knowledge. RECORDING & TRANSCRIPTION Record meetings with live audio waveform visualization so you always know the mic is active. When the meeting ends, transcription begins immediately. Choose the transcription engine that fits your workflow: - WhisperKit: on-device, no internet required, supports speaker identification - Apple Speech: system-level on-device transcription - OpenAI Whisper: cloud transcription with high accuracy - Groq: fast cloud transcription - Custom endpoint: bring your own transcription service Speaker diarization via WhisperKit and SpeakerKit identifies who said what, so your notes reflect the actual conversation rather than a wall of anonymous text. AI-POWERED NOTES After transcription, Resonance enhances your notes into a structured document: - Summary of the meeting - Key decisions made - Action items with ownership - Discussion points covered - Open questions left unresolved Chat with AI about any meeting to ask follow-up questions, clarify details, or extract specific information without scrolling through the full transcript. POWERFUL SEARCH & ORGANIZATION Cross-meeting search lets you find any topic, name, or decision across your entire meeting history in seconds. Custom vocabulary teaches Resonance your domain — add project names, acronyms, technical terms, and people so transcription and notes use the right words from day one. CHOOSE YOUR AI BACKEND Resonance supports seven LLM providers for note enhancement: - Apple Foundation Models (fully on-device) - Ollama (local server, stays on your network) - Google Gemini - OpenRouter - OpenAI - Anthropic Claude - Custom endpoint No account is required. Configure only the providers you want to use. PRIVACY FIRST The developer collects zero data. No analytics, no advertising, no account required. Your recordings, transcripts, and notes live in the app's sandboxed storage on your device. API keys are stored in the iOS Keychain and are never logged or shared. The full on-device workflow — WhisperKit for transcription and Apple Foundation Models for note enhancement — requires no internet connection and sends nothing to any server. Cloud AI services are optional and off by default; they activate only when you add an API key. iCloud sync via CloudKit keeps your meeting history in sync across your own Apple devices using your personal Apple ID. PLATFORM SUPPORT Resonance runs on iPhone, iPad, Mac, and Apple Vision Pro. The interface uses Apple Liquid Glass design, available on iOS 26 and later. ---