Press a hotkey anywhere on your system, speak naturally, and Telvr transcribes your voice in real-time using Whisper. The finished text is automatically inserted at your cursor position — no copy-paste, no app switching.

Which languages are supported?

Telvr supports 50+ languages via OpenAI's Whisper large-v3 model. Language detection is automatic — just speak in your preferred language and Telvr handles the rest.

Do I need a subscription?

No. Telvr uses a pay-as-you-go model: EUR 3 per month infrastructure fee plus EUR 0.03 per minute of usage. No lock-in, no auto-renewal. You top up your balance and use it at your own pace.

Does Telvr work offline?

Currently, Telvr requires an internet connection for cloud-based transcription via Groq. A Community Edition with local processing using your own API key is planned for the future.

Which apps are supported?

Telvr works system-wide — it inserts text at the cursor position in any application. Email clients, chat apps, code editors, browsers, word processors — if you can type in it, Telvr works there.

All data is transmitted via TLS encryption. Audio recordings are not permanently stored after transcription. Groq processes your audio under a Data Processing Agreement (DPA). We do not sell or share your data.

← Blog2026-02-20

Speech-to-Text Comparison 2026: Telvr vs Whisper vs Native Dictation

The State of Speech-to-Text in 2026

Voice input has evolved beyond simple dictation. Modern speech-to-text tools now handle multiple languages, remove filler words, and format output intelligently. But which solution actually fits into a desktop productivity workflow?

We compared three approaches: Telvr (push-to-talk with AI enrichment), OpenAI Whisper (open-source transcription), and native OS dictation (macOS Dictation / Windows Voice Typing).

Accuracy

All three solutions deliver strong baseline accuracy for English in quiet environments. The differences emerge in real-world conditions:

Telvr uses Whisper large-v3 via Groq's inference API, achieving near-identical accuracy to standalone Whisper with significantly lower latency. The AI enrichment layer corrects grammar and removes fillers automatically.
Whisper (self-hosted) provides excellent raw transcription but requires post-processing for clean output. Running locally demands significant GPU resources.
Native dictation works well for short phrases but struggles with technical terminology, mixed-language input, and longer passages.

Speed and Latency

Speed matters when voice input replaces typing in real-time workflows:

Telvr: Under 2 seconds end-to-end latency. Cloud processing via Groq's optimized inference means no local hardware requirements.
Whisper (local): Depends entirely on your hardware. A modern GPU delivers 2-5 seconds for typical passages. CPU-only can take 10-30 seconds.
Native dictation: Near-instant for short phrases. Longer passages may introduce delays and accuracy drops.

Integration

This is where the approaches diverge most:

Telvr: System-wide hotkey inserts text directly at your cursor position. Works in any application without switching windows. Six AI enrichment modes transform raw speech into emails, meeting notes, or cleaned text.
Whisper: Requires a custom pipeline. You need to record audio, run transcription, and manually paste the result. Several open-source wrappers exist, but none match system-wide integration.
Native dictation: Built into the OS but limited to supported text fields. No enrichment, no formatting, no multi-mode output.

Language Support

Telvr: 50+ languages via Whisper large-v3. Automatic language detection.
Whisper: Same model, same language support. Self-hosted gives full control.
Native dictation: Varies by OS. macOS supports ~60 languages, Windows Voice Typing is more limited.

Pricing

Telvr: EUR 3/month infrastructure + EUR 0.03/minute usage. 14-day free trial with EUR 3 starter credit.
Whisper (self-hosted): Free (open-source), but requires GPU hardware or cloud compute costs.
Whisper (API): $0.006/minute via OpenAI API.
Native dictation: Free, included with the OS.

The Verdict

Choose Telvr if you want voice input that works everywhere on your desktop without setup complexity. The AI enrichment modes turn raw speech into formatted, professional text — something neither Whisper nor native dictation offers out of the box.

Choose Whisper (self-hosted) if you need full control over your data, have capable hardware, and are comfortable building a custom pipeline.

Choose native dictation for quick, casual voice input where accuracy and formatting are not critical.

The biggest differentiator is integration depth. Telvr is the only solution that combines transcription, AI processing, and system-wide text insertion into a single hotkey. For desktop productivity, that integration eliminates the friction that makes other solutions feel like a workaround rather than a tool.