`benecaster_html_formatting_model`

Filter Premium transcription

The OpenAI Chat Completions model Format transcript uses for an episode. A non-string or empty return is ignored and gpt-4o-mini is used. It is read once per run, in the background.

The transcript is sent in parts of about 2,000 words, and the run fails, storing nothing, if any answer is cut off. The cost is the podcaster’s: a larger model costs more than the typically under $0.01 per episode the button says.

Parameters

Name	Type	Default	Description
`$model`	`string`	—	`gpt-4o-mini`
`$episode_id`	`int`	—	ID of the episode

Returns: string

Need this built rather than just documented? See our services →

Name	Set by	What it does	Kept for
`wordpress_logged_in_*`	Benecaster	Keeps you signed in to your Benecaster account.	Until you sign out, or two weeks with "Remember Me".
`wordpress_sec_*`	Benecaster	Secures your signed-in session over HTTPS.	Until you sign out, or two weeks with "Remember Me".
`wp-settings-*`	Benecaster	Remembers your admin screen preferences.	One year.
`benecaster_ls_purchase_sign_in`	Benecaster	Signs you in to your new account on the page you bought from, once your purchase has gone through.	Thirty minutes, or until you are signed in.
`benecaster_ls_cookie_consent`	Benecaster	Remembers the choice you make on this banner, so we do not ask again.	One year.

Name	Set by	What it does	Kept for
`__stripe_mid`	Stripe	Lets Stripe recognise the device across payment attempts to detect card fraud.	One year.
`__stripe_sid`	Stripe	The same fraud check, for the current visit only.	Thirty minutes.

Parameters

Related Recipes