Supported Models

Per-family recipes covering weight conversion, launch flags, and parallelism choices.

Miles ships ready-to-run recipes for every model family listed below. Each page covers weight conversion, parallelism, and the launch script in the order you’d actually run them.

Language Models

Each model name links to its recipe page.

Family	Models
DeepSeek	DeepSeek-V4 Pro DeepSeek-V4 Flash DeepSeek-V3.2 DeepSeek-V3
Thinking Machines	Inkling Inkling-Small
Qwen	Qwen3.8-Flash-Next Qwen3.8-27B Qwen3.6 MoE Qwen3.6 Qwen3.5-35B-A3B Qwen3.5-4B / 9B / 27B Qwen3-Next-80B-A3B-Thinking Qwen3-30B-A3B / 235B-A22B Qwen3-0.6B / 1.7B / 4B / 8B / 14B / 32B
GLM	GLM-5.3-Flash GLM-5.2 GLM-5.1 GLM-5 GLM-4.7-Flash GLM-4.5
Kimi	Kimi-K3 Kimi-K2.6 Kimi-K2.5 Kimi-K2-Instruct / Thinking
Nemotron	Nemotron-3-Ultra-550B-A55B Nemotron-3-Super-120B-A12B-FP8 Nemotron-3-Nano MoE Nemotron-3-Nano
Gemma	Gemma-4 26B-A4B Gemma-4 31B
JoyAI	JoyAI-LLM-Flash
GPT-OSS	gpt-oss-20b

Diffusion

Family	Models
Stable Diffusion	SD3.5
Qwen-Image	Qwen-Image
Wan	Wan2.2-T2V-A14B
LTX	LTX-2.3
Cosmos	Cosmos3-Nano
MiniMax	MiniMax H3

How a recipe is structured

Every recipe page follows the same six sections:

1. Model Introduction — what the model is and why miles supports it.
2. Supported Variants — model sizes + HF links.
3. Environment Setup — env vars, downloads, and HF → Megatron conversion.
4. Launch — the scripts/run_<family>.py invocation.
5. Recipe Configuration — parallelism, algorithm, rollout/SGLang, optimizer.
6. Pairs Well With — links to the advanced features that complement this recipe.

Adding a new model

Miles’s plugin architecture lets you wrap a HuggingFace implementation as a Megatron module without patching Megatron core. See Backends Beyond Megatron for the workflow.