Start free. Upgrade when you need more runs and access to larger models.
For individuals exploring model fine-tuning.
For teams and power users who need more capacity.
A run is a single end-to-end cycle: compile your behaviour spec into training data, optionally augment examples with AI, fine-tune the model, and auto-evaluate. Each run counts toward your monthly limit regardless of outcome.
The Free tier includes Qwen models for fine-tuning (Qwen 2.5 7B, Qwen 2.5 72B, and Qwen 3.5 35B-A3B). The Pro tier unlocks all base models including Llama and Mistral families in addition to all Qwen models.
Yes. You can downgrade back to the Free tier at any time. Your existing fine-tuned models and behaviour specs will remain accessible.
No. Fine-tuning and inference are handled through the Tuned Tensor platform. You only need a Tuned Tensor account to get started.
Create your free account and start fine-tuning in minutes.