Supported Models for Reinforcement Learning and Supervised Fine-Tuning

Provider	Models	Notes
Qwen 3	0.6B, 1.7B, 4B, 8B, 14B, 32B	Default for demos; supports tool-calling.
Qwen 3 (Advanced)	4B-2507, 30B-A3B, 235B-A22B, 480B-A35B	Instruct/Thinking/MoE variants; 235B/480B require sharding.
Qwen 3 Coder	30B-A3B, 480B-A35B	Code-specialized; large sizes require sharding.

Large models (235B/480B) must be sharded across multiple GPUs for inference and training.