Select model
Gemma, Qwen, Mistral...
Upload data
CSV, JSON, JSONL
Download model
Ready to deploy
Built for your use case
Fine-tune models for any domain or task
Three steps to your custom model
No coding required. No ML expertise needed.
Choose a base model
Select from Gemma, Qwen, Mistral, and other popular open-source models. Pick the size that fits your needs.
Upload your data
Upload CSV, JSON, or JSONL files with your training examples. We'll automatically detect the format.
Download & deploy
Download your model or deploy to Hugging Face with one click. Ready to use anywhere.
Why TinyTune?
Everything you need to create custom AI models
Fast training
Using LoRA, your models are ready in minutes instead of hours or days.
Data privacy
Your data is never shared or used to train other models. Fully private.
Full ownership
Download your model and deploy it anywhere. No vendor lock-in.
No expiration
Credits never expire. Use them whenever you're ready.
One-click deploy
Push to Hugging Face with one click, or download and run anywhere you want.
Open models
Fine-tune popular open-source models with commercial-friendly licenses.
Simple, transparent pricing
Pay once, use anytime. No subscriptions, no hidden fees.
One-time purchaseCredits never expireNo recurring charges
40+ models to choose from
Fine-tune popular open-source LLMs from leading providers
Frequently asked questions
Everything you need to know about TinyTune