Overview
Unsloth's mission is to make AI as accurate and accessible as possible. Train, run, evaluate and save gpt-oss, Llama, DeepSeek, TTS, Qwen, Mistral, Gemma LLMs 2x faster with 70% less VRAM. With Koyeb, you can deploy Unsloth's Docker image and start model training in seconds.
Get up to $200 in credit to get started!
Try it out
Choose a value for JUPYTER_PASSWORD in the Environment variables section of the Koyeb control panel for your Service. This password is how you access the environment through HTTPS. To access the environment through SSH, provide a value for SSH_KEY.
Once the Unsloth server is deployed, you can start training with Unsloth through the notebooks included in the image or by adding your own. Access the Instance using your Koyeb App URL, which is similar to: https://<YOUR_APP_NAME>-<YOUR_KOYEB_ORG>.koyeb.app.
The default Instance uses an NVIDIA L40 GPU, but you can change the configuration to suit your needs.
