WhisperDirect

WhisperDirect is a high-accuracy speech-to-text and summarization app that works with your own API key. No subscription required — you only pay OpenAI’s usage fees when you need it, making it more cost-effective than subscription-based apps. Pricing & Trial • Free trial: 5 sessions included • After the trial: one-time in-app purchase unlocks unlimited use of current features • API usage billed directly by OpenAI (the app does not charge for API usage) Cost Guide • With $5, you can transcribe about 14 hours of audio • Whisper API ≈ $0.006 per minute (≈ $0.36 per hour) • OpenAI API pricing → https://openai.com/ja-JP/api/pricing/ Models for summaries and meeting minutes Choose from compact, low-cost models: • GPT-4.1-nano • GPT-4.1-mini • GPT-5-nano • GPT-5-mini Even long texts (1,000–2,000 words) can usually be processed for just a few cents per run. Main Features • Record with the microphone button and instantly convert to text • Import audio files (or directly from the share sheet) • Import video files (audio extracted and compressed automatically) • Playback-synced highlighting of transcript segments • Insert timeline markers (configurable in 5-second steps) • Generate summaries and meeting minutes (prompts editable in Settings) • OCR transcription from images (supports multiple images, all processed locally with no extra API cost) • Export audio, text, summaries, minutes, or subtitles (VTT / SRT) • Automatically post transcripts/summaries/minutes to Slack • Estimate costs in Settings (based on audio length and character count) • Other customization options (LLM model, timeline interval, prompts, etc.) 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 • An API key (such as OpenAI) is required • Pricing and available models may change according to OpenAI’s offerings