Configuration File

whispercom stores your provider and model preference in a JSON file:

~/.config/whispercom/config.json

Schema

{
  "provider": "custom",
  "model": "openai/gpt-4o",
  "apiKey": "sk-or-v1-...",
  "baseUrl": "https://openrouter.ai/api/v1"
}

Field	Type	Description
provider	string	One of: anthropic, openai, google, mistral, groq, custom
model	string	A model ID valid for the selected provider
apiKey	string	(Optional) The API key for the provider. If omitted, the environment variable is used.
baseUrl	string	(Optional) The base URL, only used if the provider is set to custom.

Editing Manually

You can edit the file directly in any text editor instead of using whis configure.

# macOS / Linux
nano ~/.config/whispercom/config.json

# Windows
notepad $env:USERPROFILE\.config\whispercom\config.json

Resetting Configuration

Delete the config file to force a fresh setup on the next whis run:

rm ~/.config/whispercom/config.json

Valid Provider / Model Combinations

provider	Valid model values
anthropic	claude-opus-4-5, claude-sonnet-4-5, claude-haiku-4-5, or any typed ID
openai	gpt-4o, gpt-4o-mini, o3-mini, or any typed ID
google	gemini-2.5-pro-exp-03-25, gemini-2.0-flash, gemini-2.0-flash-thinking-exp, or any typed ID
mistral	mistral-large-latest, mistral-small-latest, codestral-latest, or any typed ID
groq	llama-3.3-70b-versatile, llama-3.1-8b-instant, deepseek-r1-distill-llama-70b, or any typed ID
custom	Any OpenAI-compatible model ID (no presets)

WARNING

If you specify an invalid model ID, whispercom will pass it directly to the provider API and you'll see an error from the provider. Run whis configure to pick a valid model from the list.