models: array of object {id, display_name, model_size, 4 more}
id: string
Identifier to pass as model when creating a job.
display_name: string
Human-readable model name.
model_size: optional string
Parameter count label (e.g. “8B”, “70B”).
context_length: optional number
Maximum context window in tokens.
methods: array of string
Training methods supported (e.g. [“sft”, “dpo”]).
training_types: array of string
Training types supported (e.g. [“lora”, “full”]).
train_with_reasoning: boolean
Whether this model trains on reasoning traces. When false, mapping a column_mapping.reasoning_trace on this model is rejected rather than trained without the trace.
