> ## Documentation Index
> Fetch the complete documentation index at: https://staging.docs.trywhitney.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Supported models

> The provider and model pairs currently admitted by Whitney.

Use `GET /v1/training/capabilities` as the source of truth at runtime. Every
returned row is supported for use through the operations it advertises.

| Provider | Models                                                                                         |
| -------- | ---------------------------------------------------------------------------------------------- |
| Modal    | Qwen/Qwen3.5-0.8B, Qwen/Qwen3.5-4B, Qwen/Qwen3.5-9B                                            |
| Tinker   | Qwen/Qwen3.5-4B, Qwen/Qwen3.5-9B, Qwen/Qwen3.5-9B-Base, Qwen/Qwen3.8-27B, moonshotai/Kimi-K2.6 |

Prefer `provider: auto` for application code. Choose `modal` or `tinker` only
when reproducibility, provider credentials, or a provider-specific resource
profile requires it.

Model inclusion is admission, not a promise that different providers produce
identical numerics, throughput, checkpoint formats, or loss curves. Inspect
each model row's `operations`, `algorithms`, `checkpoint_modes`, and
`artifact_modes` before allocating work.