No reviews yetBe the first to leave a review for Nanik — Bedtime Stories for Kids
Maker
📌
Ever tried being a wildly creative parent when you're running on 1% battery?
I loved inventing wild bedtime stories for my daughter, like a pencil in debt to an elephant at the bank. But late nights as a product manager often left me too drained to even speak, let alone create new heroes on the spot. After a recent layoff, I turned that exact parenting struggle into my next project. That’s how Nanik was born.
Here is what makes Nanik special.
- Age-Aware Customization. You completely control the hero, the world, and the specific challenge, while the model automatically adapts the vocabulary and complexity to respect your child's exact age.
- Write prompts to teach real-world concepts like "teach her what AI is" or "help her save money for a bicycle."
- Advanced Voice Cloning & Translation. You only need a 10sec single voice clone to generate stories in 100+ languages. Crucially, the model preserves your natural, comforting tone instead of turning you into a weird, generic "radio announcer" like many popular voice closing services do, so your child always recognizes that it's actually you telling the story.
- Bring Toys to Life - Generate the story's visual images directly from photos of your child's actual favorite toys.
- Episodic Adventures- Keep the magic going night after night. You can seamlessly continue any story with a brand new chapter, which is perfect for building daily learning habits.
Under the hood, Nanik runs on an amazingly capable, custom-trained model designed specifically to weave therapeutic and educational narratives for kids. To make sure Nanik’s stories were genuinely helpful and engaging, we didn't just rely on standard AI. We interviewed child therapists to understand how to help kids process their days, and we studied the narrative structures of legendary national tale writers across the globe, from Hovhannes Tumanyan (Armenia) and Gianni Rodari (Italy), to Charles Perrault (France) and Maurice Sendak (US).
🚀 Try it today it freeeeeee
Nanik is already submitted and waiting for official App Store review, but you are completely free to bypass the wait! Join our TestFlight testing right now to enjoy full functionality today.
Report
@gor_kroyan1 Great idea!Keep going you’re doing a great job.😍
Report
the voice cloning + photos-of-your-kid combination is where I'd slow down before trying this. that's biometric data on a minor, not just a normal app upload. what's the retention policy on the voice sample and the toy/child photos once a story's generated - deleted after processing, or kept on a server to make future stories faster? and is anything used to train or fine-tune the model further, or is it strictly per-user inference?
Report
Maker
Hi @galdayan, thanks for the raising this up. And
1. Parents Only, Not for Kids The app is designed exclusively for parents and is not rated as an app for children. The camera is meant strictly for snapping photos of toys, not kids. As a consideration we already researched a local model which in case of taking kids photo rejects to send. thanks for bringing this up
2.No Voice Retention. The raw audio sample is deleted the moment the voice clone is created. From then on, only the secure voice reference is used, never the original recording.
3. Zero Model Training: We do not use any photos or voice data to train or fine-tune our models. Everything runs via per-user inference on our own production-ready servers.
Report
@gor_kroyan1 appreciate the detailed breakdown, that covers exactly what I was worried about. the local reject-model for kid photos is a nice touch, wasn't expecting that
Report
Love the idea! Most apps like this usually make you sound like a robot narrator, so hoping this one is different. How long does it take to make one story? Would love to test.
Report
Maker
@margarita_kripulevich The story generation takes 30 sec and voiceover takes roughly 1 minutes. (Besides recording your 10 sec voice for cloning)
Report
the voice cloning from a 10s clip is genuinely clever execution, but "therapeutic" is a pretty loaded word to put on the tin. interviewing child therapists during development is different from having a clinician actually review or sign off on the stories a given child gets, especially ones triggered by a parent's own prompt like a tough day at school. is there any review layer on the generated output, or is it fully unsupervised once the prompt goes in?
Report
My daughter's face when she heard my voice narrating a story about us was priceless. Love that parents can actually shape the content too.
Report
Maker
@ariflq7p this words are so important for me. thanks
tested yesterday , my kids were excited :) congrats with the launch !
Report
Maker
@arsen_am Thanks Arsen. This words are important for me to continue improving
Report
Congratulations! Great app. It might be worth adding a note in the description reminding parents not to over-rely on it and to keep telling their own stories too :) Awesome job, love it!
@gor_kroyan1 Great idea!Keep going you’re doing a great job.😍
the voice cloning + photos-of-your-kid combination is where I'd slow down before trying this. that's biometric data on a minor, not just a normal app upload. what's the retention policy on the voice sample and the toy/child photos once a story's generated - deleted after processing, or kept on a server to make future stories faster? and is anything used to train or fine-tune the model further, or is it strictly per-user inference?
Hi @galdayan, thanks for the raising this up. And
1. Parents Only, Not for Kids The app is designed exclusively for parents and is not rated as an app for children. The camera is meant strictly for snapping photos of toys, not kids. As a consideration we already researched a local model which in case of taking kids photo rejects to send. thanks for bringing this up
2.No Voice Retention. The raw audio sample is deleted the moment the voice clone is created. From then on, only the secure voice reference is used, never the original recording.
3. Zero Model Training: We do not use any photos or voice data to train or fine-tune our models. Everything runs via per-user inference on our own production-ready servers.
@gor_kroyan1 appreciate the detailed breakdown, that covers exactly what I was worried about. the local reject-model for kid photos is a nice touch, wasn't expecting that
Love the idea! Most apps like this usually make you sound like a robot narrator, so hoping this one is different. How long does it take to make one story? Would love to test.
@margarita_kripulevich The story generation takes 30 sec and voiceover takes roughly 1 minutes. (Besides recording your 10 sec voice for cloning)
the voice cloning from a 10s clip is genuinely clever execution, but "therapeutic" is a pretty loaded word to put on the tin. interviewing child therapists during development is different from having a clinician actually review or sign off on the stories a given child gets, especially ones triggered by a parent's own prompt like a tough day at school. is there any review layer on the generated output, or is it fully unsupervised once the prompt goes in?
My daughter's face when she heard my voice narrating a story about us was priceless. Love that parents can actually shape the content too.
@ariflq7p this words are so important for me. thanks
Async
tested yesterday , my kids were excited :) congrats with the launch !
@arsen_am Thanks Arsen. This words are important for me to continue improving
Congratulations! Great app. It might be worth adding a note in the description reminding parents not to over-rely on it and to keep telling their own stories too :) Awesome job, love it!
@vahe_ghazaryan1 Vahe jan exactly. Have in my list