Together AI launches Together Post-Training for custom open-model RL and SFT
Article timeline
The articles in this story. Times are ET.
Together AI@togethercomputeOfficialTogether AI launches Together Post-Training for custom open-model trainingAITogether AI says its Together Post-Training service is live, letting users turn open models into custom models they own using supervised fine-tuning (SFT) or reinforcement learning (RL). Users define the data, rewards, environments and training loop, while Together runs the underlying infrastructure.
- Together AI@togethercomputeOfficialTogether AI launches Together Post-Training for custom model RL and SFT
AITogether AI introduces Together Post-Training, a service for building custom models from environment rewards via RL and supervision data via SFT. Users define the training loop while Together manages the underlying infrastructure, with LoRA and full-weight training supported. The service launches with open-weight models including Muse Spark Open, GLM 5.3, Kimi K3, and Qwen 3.8-27B, and is offered at $9 per GPU-hour of B200 usage for the first six weeks.