What is new in Decklyn.
We ship improvements regularly. Here is what has changed recently, newest first.
Recent releases.
Brand kit for Pro and Team
Pro and Team subscribers can now set a brand color, a preferred narration voice, and a logo in their account settings. Decklyn uses these on every deck they make, so the generated visuals lean on the brand color and the narration sounds consistent across the team.
- Brand color is applied to the generated visual for every slide, in one consistent style.
- Set a preferred narration voice once in brand settings and Decklyn uses it automatically.
- Team plan adds a shared brand kit so every team member makes decks that look and sound the same.
Eleven new narration voices and per-slide regeneration
Pro and Team now have access to eleven additional ElevenLabs narration voices, covering a wider range of accents and delivery styles. All plans can now regenerate just one slide's visual or narration without rebuilding the whole deck.
- New voices include US English, British English, Australian English, and a clear neutral voice optimized for technical topics.
- Per-slide regeneration: click the refresh button on any slide to replace its image or narration independently.
- Free plan keeps the default voice; Pro and Team unlock the full voice library.
Faster generation and structured JSON pipeline
Generation is now roughly 40% faster end to end. The language model now returns a strict validated JSON schema for the whole deck, which means fewer retries and more consistent slide quality. Image generation and narration now run in parallel after the structure is confirmed.
- Deck outline, slide copy, and narration script are now generated in one structured call with Zod validation.
- Slide images and narration audio generate in parallel, cutting the wait for a complete deck by roughly half.
- Edge cases that previously caused a generation to hang now fail fast with a clear error and a retry option.