Introducing Quorlyn Snap 2.0

Quorlyn Snap 2.0 is Quorlyn’s most capable speech-to-speech voice model, with improved intelligence, transcription accuracy, and conversational capabilities.

Try Snap 2.0

View the voice docs

Speech-to-speech model card for Quorlyn Snap 2.0 with 1.5–2× word error rate versus dedicated STT, ~60% fewer reasoning tokens than 1.0, and $0.08 per minute of audio

What’s new

More accurate transcription

1.5–2× better word error rate than dedicated STT models across 24 languages. Up to ~10× better in noisy, telephony conditions.

Efficient parallel reasoning

Reasons while speaking with ~60% fewer reasoning tokens than 1.0, so tool calls and responses feel snappier.

More natural conversation

Shorter sentences, one question at a time, less filler—trained to sound like real conversation while still driving complex workflows.

quorlyn-voice-latest switches over on August 5

On August 5, 2026, quorlyn-voice-latest will move from snap-1.0 to snap-2.0. No action needed to upgrade. To stay on 1.0, pin snap-1.0 before then.

Priced at $0.08/min of audio. Full details in the announcement and pricing docs.

Try Snap 2.0

Read the announcement

See voice pricing

Privacy · Terms · Unsubscribe