Bonfire Terminal vs @capacitor-mlkit/genai-speech-recognition
@capacitor-mlkit/genai-speech-recognition is a free, open-source, on-device speech SDK using Google ML Kit for fully private, offline mobile transcription, earning a high 81 sovereign score. Bonfire Terminal brings local Whisper speech into a complete on-device agent, with messaging bridges and BYO-key models, delivered as a one-time perpetual license.
Bonfire Terminal vs @capacitor-mlkit/genai-speech-recognition: the 16-factor comparison
| Rating factor | @capacitor-mlkit/genai-speech-recognition | Bonfire Terminal |
|---|---|---|
| Category | Voice AI / Speech | AI Terminal / Local Agent |
| Popularity | Not disclosed | Growing · indie |
| Hosting architecture | On-Device Local | On-Device Local |
| Licensing model | Open Source | One-Time Perpetual |
| Data / training policy | On-device ML Kit fully private | Fully local — nothing leaves device |
| Offline capability | Yes | Yes |
| Pricing tier range | Free (OSS) | One-time license |
| Execution unit cost | Unlimited Local Compute | Unlimited Local Compute |
| API-key markup | N/A | No — BYO key at cost |
| Supported LLM providers | Not disclosed | OpenAI; Anthropic; Google; Local; Any |
| Auth / permission types | Local/None | API Key; Local/None |
| Phone / messaging bridges | None | Telegram; WhatsApp |
| Speech recognition engine | Google ML Kit (on-device) | Whisper (local) |
| Local media processing | Partial | Yes |
| Sovereign score | 81 / 100 | 93 / 100 |
The sovereignty gap
Where @capacitor-mlkit/genai-speech-recognition lands on each weighted pillar (its bar) — the red marker is Bonfire's score. Auto-derived from the 16 factors above.
Which should you choose?
Choose @capacitor-mlkit/genai-speech-recognition if…
Choose @capacitor-mlkit/genai-speech-recognition if you build Capacitor mobile apps and want free, private, offline on-device speech via Google ML Kit.
Choose Bonfire Terminal if…
Choose Bonfire Terminal if you want on-device Whisper speech inside a full agent with bridges and BYO-key models, ready to run.
Bonfire Terminal vs @capacitor-mlkit/genai-speech-recognition: FAQ
Is @capacitor-mlkit/genai-speech-recognition private?
Yes. It uses on-device Google ML Kit, so transcription is fully private and never leaves the device. Bonfire Terminal matches this with local Whisper, keeping audio, transcripts, and agent activity on your machine rather than any cloud.
Is Bonfire Terminal cheaper than @capacitor-mlkit/genai-speech-recognition?
The Capacitor ML Kit SDK is free and open source, so it wins on raw price for developers. Bonfire Terminal charges a one-time license but delivers a finished on-device agent, not just a mobile speech component you must integrate into an app.
Can Bonfire Terminal replace @capacitor-mlkit/genai-speech-recognition?
For end users needing transcription, yes; for mobile developers, no. The ML Kit SDK is a Capacitor plugin for building apps, while Bonfire Terminal is a ready-made on-device agent. They target different audiences around the same on-device speech goal.
What can Bonfire Terminal do that @capacitor-mlkit/genai-speech-recognition can't?
Bonfire Terminal is a complete on-device agent with local Whisper, local media processing, Telegram and WhatsApp bridges, and BYO-key LLM access. The ML Kit SDK only provides on-device speech recognition for Capacitor apps, with no agent or model features.
How do I switch from @capacitor-mlkit/genai-speech-recognition to Bonfire Terminal?
If you only needed private on-device transcription rather than to build a mobile app, install Bonfire Terminal and use its local Whisper engine directly, avoiding any Capacitor integration work.
Does Bonfire Terminal work offline vs @capacitor-mlkit/genai-speech-recognition?
Both run fully offline. The ML Kit SDK transcribes on-device without a connection, and Bonfire Terminal likewise runs Whisper and media processing offline, only reaching the network when you invoke a hosted model.