Documentation Index
Fetch the complete documentation index at: https://docs.auditynow.com/llms.txt
Use this file to discover all available pages before exploring further.
Audity lets you choose which AI model powers your audits, deliverables, and research. You can change your default model at any time from Dashboard → Settings → AI (/dashboard/settings/ai).
Supported providers
| Provider | Model | Role |
|---|
| Anthropic | Claude (latest) | Primary default |
| OpenAI | GPT-4o | Fallback / alternative |
| Google | Gemini | Alternative |
| Mistral | Mistral (latest) | Alternative |
| DeepSeek | DeepSeek (latest) | Alternative |
Choosing a model
- Go to Dashboard → Settings → AI.
- Select your preferred provider and model from the dropdown.
- Click Save.
All subsequent AI operations, document analysis, interview analysis, stakeholder memo generation, questionnaire generation, and web research summaries, use the selected model.
Provider tradeoffs
Anthropic Claude (primary default)
Claude is Audity’s primary model. It performs well on long-document analysis, nuanced consulting deliverables, and structured output. Recommended for most consulting teams.
OpenAI GPT-4o
GPT-4o is the recommended fallback. It is well-suited for structured JSON output and broad general knowledge. A good alternative if Claude is unavailable or you prefer OpenAI’s output style.
Google Gemini
Gemini offers strong performance on tasks with long context windows and multimodal content. Consider Gemini if your audits involve large document sets.
Mistral
Mistral models are efficient and cost-effective for straightforward analysis tasks. A good choice if you want to conserve credits on lighter workloads.
DeepSeek
DeepSeek provides competitive performance on analytical and reasoning tasks. Suitable as an alternative if you have access and prefer its output characteristics.
Credit usage and model choice
All supported models draw from the same credit pool. Credit costs are based on the volume of content processed, not the specific model selected. See Credits and usage for more detail.
