Brand Voice

The Brand Voice service lets you create, manage, and apply brand voices to ensure consistent tone and style across all generated content.

Concepts

Brand Voices

A brand voice defines how your content should sound — including tone, vocabulary, formatting rules, and writing style. Brand voices can be:

• Created manually with detailed style settings
• Imported from URLs, documents, or text descriptions using AI analysis
• Duplicated from existing brand voices
• Upgraded for enhanced quality using AI

Brands

Brands represent the company or product associated with a brand voice. They contain contact information, social media links, and brand values.

DS Prompt

Each brand voice can generate a "DS prompt" — a formatted system prompt that can be passed to the content generation API to apply the brand voice to generated text.

Endpoints

All endpoints use the base path: https://app.neuroflash.com/api/brand-voice-service

Brand Voices

Method	Endpoint	Description
POST	/v1/workspaces/{id}/brand-voices	Create a brand voice
GET	/v1/workspaces/{id}/brand-voices	List brand voices
GET	/v1/workspaces/{id}/brand-voices/{id}	Get a brand voice
PUT	/v1/workspaces/{id}/brand-voices/{id}	Update a brand voice
DELETE	/v1/workspaces/{id}/brand-voices/{id}	Delete a brand voice
POST	/v1/workspaces/{id}/brand-voices/{id}/duplicate	Duplicate a brand voice
POST	/v1/workspaces/{id}/brand-voices/{id}/ds-prompt	Get DS prompt for content generation
POST	/v1/workspaces/{id}/brand-voices/{id}/upgrade	Upgrade brand voice quality

Brands

Method	Endpoint	Description
POST	/v1/workspaces/{id}/brands	Create a brand
GET	/v1/workspaces/{id}/brands	List brands
GET	/v1/workspaces/{id}/brands/{id}	Get a brand
PUT	/v1/workspaces/{id}/brands/{id}	Update a brand
DELETE	/v1/workspaces/{id}/brands/{id}	Delete a brand

Imports

Imports return a draft — a second API call is required to save it

The three import endpoints analyse your content and return a brand voice proposal. No data is written to your workspace until you call POST /brand-voices with the extracted fields. See the Persisting an Import Draft section for the complete two-step workflow.

Method	Endpoint	Description
POST	/v1/workspaces/{id}/brand-voice-text-imports	Import from text
POST	/v1/workspaces/{id}/brand-voice-url-imports	Import from URL
POST	/v1/workspaces/{id}/brand-voice-file-imports	Import from file

Importing Brand Voices

All three import endpoints (URL, text, file) share three required parameters beyond workspaceId that are easy to get wrong:

Parameter	Type	Description
country	string	2-letter lowercase country code (e.g. us, de, gb)
language	string	2-letter lowercase language code (e.g. en, de, fr)
audienceModelId	string	Audience embedding model ID — use the default value below if unsure

Default audienceModelId: 538b1efc6f88ad88feebf7acd8c618facb54fe82

Allowed locale pairs

country and language must form a valid pair. The API validates this server-side and returns a 400 error if the combination is not allowed. Supported pairs:

country	Valid language values
us	en
gb	en
uk	en
de	de
at	de
ch	de, fr
fr	fr
es	es, ca
mx	es
it	it
nl	nl
pl	pl
pt	pt
br	pt
se	sv
tr	tr
hu	hu
hr	hr
cz	cs
cn	zh

Safe defaults: use country: "us", language: "en" for English content or country: "de", language: "de" for German content.

Example: Import from URL

curl -X POST "https://app.neuroflash.com/api/brand-voice-service/v1/workspaces/{workspace_id}/brand-voice-url-imports" \
  -H "Authorization: Bearer YOUR_ACCESS_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "url": "https://your-brand.com",
    "country": "us",
    "language": "en",
    "audienceModelId": "538b1efc6f88ad88feebf7acd8c618facb54fe82"
  }'