Brand Voice

One brand library. Every chat on-context.

Save your company facts and product details once. Pick a brand and product in any chat, and AresGen prepends the context to your prompt across 12 routed models.

Build your brand library See it in AI Chat

Brand picker (in any chat form)

Brand

Ledgerly

Product

Audit Log 2

Stored in your accountSynced to every routed modelREST API for external appsFree plan includes 1 Brand

backend surfaces

Library · Chat injection · REST API

chat applications

AI Chat · AIChatPro · Workbook · Social Media Agent

routed AI models

Same rail as /tools/ai-chat and /tools/ai-code

public REST endpoint

GET/POST/DELETE /api/brandvoice

Three backend surfaces.

Each surface ships a focused responsibility. Together they cover how Brands enter, apply, and leave AresGen.

Brand library

Create one or more Brands. Each Brand holds a company profile plus a list of products or services.

Chat injection

Pick a Brand and a Product in any chat. AresGen prepends a structured company-and-product paragraph before your prompt runs.

REST API

GET, POST, and DELETE Brands programmatically. Useful when an external CMS or CRM owns the source of truth.

Four chat surfaces, one Brand picker.

Surface #2 (chat injection) fires in every chat path that runs through GeneratorController.

AI Chat

Default chat panel: picker in the chat form below the model selector.

AIChatPro

Document, code, and image chat all read the same Brand picker.

Workbook

Long-form drafting surface inherits the AIChatPro brand context.

Social Media Agent

Brand context is appended before the agent system prompt resolves.

What lives inside a Brand.

Two simple records. Company describes who you are. Product describes what you sell.

Company

Namename
Industryindustry
Descriptiondescription
Websitewebsite
Taglinetagline
Logologo
Brand colourbrand_color
Specific instructionsspecific_instructions

Product

Product namename
Type (Service or Product)type
Descriptiondescription
Key featureskey_features

See the prompt with and without Brand context.

Pick an example. Left column is the raw prompt you typed. Right column is what the routed model actually receives.

Without Brand

Write a launch post for our new audit log feature.

With Brand

Focus on my company and Product's information:
The company's name is Ledgerly. The company is in the compliance-tech industry. The company's tagline is "Evidence on autopilot." The Product's name is Audit Log 2. The Product's key features are tamper-proof timeline, SOC2 export.

Write a launch post for our new audit log feature.

Works with all 12 routed models.

The injected paragraph travels with your prompt to whichever model you pick. Same library, same context.

Anthropic·claude-opus-4-7Anthropic·claude-sonnet-4-6Anthropic·claude-haiku-4-5OpenAI·gpt-5OpenAI·gpt-5-codexOpenAI·gpt-4-1Google·gemini-pro-2-5Google·gemini-flash-2-5DeepSeek·deepseek-v3DeepSeek·deepseek-r1Meta·llama-3-3-70bMistral·mistral-large-2

Where Brand Voice earns its keep.

B2B SaaS launch in your house style

Pick "Ledgerly" + "Audit Log 2"; every launch asset reads like the brand instead of like a generic chatbot.

On-brand customer support replies

Specific-instructions guide tone; product key-features keep refunds, returns, and shipping language consistent.

One workspace, every client

Agencies keep one Brand per client. Switch brands per project, with no separate accounts and no copy-paste briefs.

Bilingual reply, single source of truth

Write the prompt in any language the model supports; your brand facts arrive in the same chat context regardless.

Included on every paid plan.

Free plan includes one Brand and unlimited Products under it. Paid plans lift the Brand count.

See pricing

FAQ

What does Brand Voice actually do?

When you pick a Brand and a Product in a chat, AresGen builds a paragraph of your company name, industry, tagline, description, specific instructions, and the selected product's name and key features. That paragraph is injected as a user-role message into chat history before your prompt is sent to the routed model.

Do I upload writing samples to train a voice?

No. There is no sample-based training step. You fill structured fields (name, industry, description, tagline, brand colour, specific instructions, products) and that information is injected as a user-role message into chat history on demand. Editing a field takes effect on the very next message.

Which languages does Brand Voice cover?

Brand Voice is language-agnostic. The field values are stored exactly as you type them, and the routed model handles the prompt in whichever language you and your model support. We do not publish a language list because no language gating happens in the brand layer itself.

Does Brand Voice apply to the AI Writer templates?

Not today. Brand Voice fires inside the chat pipeline: AI Chat, AIChatPro (Document, Code, Image), Workbook, and the Social Media Agent. The AI Writer template generators run a different path and do not currently consume the Brand library.

Can I keep multiple brands in one account?

Yes. Each Brand sits under your user account and you can switch between them inside the chat picker. Agencies and holding companies use this to keep separate clients in one workspace.

Can external apps read or write Brands?

Yes. The /api/brandvoice surface supports GET (list and single), POST (create), and DELETE. Authenticate with an API token the same way you would for any other AresGen API endpoint.

AI Chat

The canonical chat surface where Brand Voice picker lives in the chat form.

/tools/ai-chat →

AI Code

Code-aware chat under AIChatPro: same Brand picker, same injection mechanism.

/tools/ai-code →