Your AI-Powered
Voice Assistant

TokKong runs all AI models directly on your device using Apple's Neural Engine and Core ML. The Whisper model for transcription and various LLMs (like Gemma, Phi-3, and Llama) are downloaded once and run locally. This means no internet connection is required after the initial setup, and your data never leaves your device.

TokKong supports over 99 languages through the Whisper AI model. This includes major languages like English, Spanish, Mandarin, French, German, Japanese, Korean, Arabic, and many more. The model automatically detects the language being spoken, or you can manually select it for better accuracy.

Absolutely. TokKong is designed with privacy as a core principle. All transcription, AI processing, and data storage happen entirely on your device. We never upload your audio, transcripts, or any personal data to cloud servers. Your conversations stay completely private.

TokKong uses OpenAI's Whisper model for speech-to-text transcription, which is renowned for its accuracy. For AI chat and text enhancement, we offer multiple models including Gemma (Google), Phi-3 (Microsoft), and Llama (Meta). You can download only the models you need and switch between them freely.

TokKong supports multiple export formats: plain text (TXT), SRT subtitles for video editing, JSON for developers, and direct sharing to other apps. You can copy to clipboard, save to Files, or share via the iOS share sheet to apps like Notes, Mail, or messaging apps.

TokKong is optimized for iPhone and iPad running iOS 16 or later. It leverages Apple's Neural Engine for maximum AI performance, so newer devices (iPhone 12 and later, iPad Pro M-series) will see faster processing. Mac support via Apple Silicon is available with macOS 14+.

TokKong supports all major audio formats including MP3, M4A, WAV, AAC, FLAC, and AIFF. You can import audio from Voice Memos, WhatsApp voice messages, the Files app, or record directly in TokKong. There's no file size limit - process hour-long recordings with ease.

Using the Whisper AI model, TokKong achieves professional-grade accuracy comparable to human transcriptionists for clear audio. Accuracy depends on audio quality, background noise, and accents. For optimal results, we recommend recording in quiet environments with the microphone close to the speaker.

Yes! TokKong fully integrates with iOS Shortcuts. You can create automations to transcribe audio files, process text with AI, or trigger recordings. This enables powerful workflows like automatically transcribing voice memos or summarizing meeting recordings.

TokKong offers a free tier with 5 transcriptions per day. For unlimited usage, you can subscribe to Pro ($9.99/month) or get Lifetime access ($79.99 one-time). All purchases are handled through the App Store with easy cancellation anytime.