Create a v2 parsing job

Parsing parameters.

Webhook configuration.

OpenAI-compatible BYOK credentials (flat root + optional channel overrides).

Happy path (one multimodal model for both channels)::

{"api_key": "...", "model": "gpt-4o", "base_url": "https://api.openai.com/v1"}

Same endpoint, different models per channel::

{
  "api_key": "...",
  "base_url": "https://api.openai.com/v1",
  "models": {"text": "gpt-4o-mini", "vision": "gpt-4o"}
}

Different endpoints per channel::

{
  "text": {"api_key": "...", "model": "...", "base_url": "..."},
  "vision": {"api_key": "...", "model": "...", "base_url": "..."}
}

Semantics:

• root api_key + base_url with model and/or models -> shared auth
• models.<channel> wins over root model for that channel
• text / vision objects fully replace the root for that channel
• a channel with no resolved model keeps server defaults