avidia contentHelp & Documentation
Back to App

AI Settings

Configure AI providers, prompts, and generation settings

AI Configuration Hierarchy

AI settings in avidia content follow a cascading hierarchy. Settings at each level override the previous level, allowing for flexible configuration from organization-wide defaults to individual flow customizations.

Settings Hierarchy

1.

Global

System-wide defaults

Overrides ↓
2.

Organization

Organization overrides

Overrides ↓
3.

Project

Project-specific settings

Overrides ↓
4.

FLOW

Individual flow settings

Available Settings

provider
AI provider (OpenAI or Anthropic)
model
Model selection (GPT, Claude, etc.)
temperature
Creativity level (0-1)
maxTokens
Maximum output length
systemPrompt
Custom system instructions
rules
Content guidelines to follow
negativePrompts
Things to avoid
glossary
Term definitions and translations
toneOfVoice
Writing tone (formal, casual, etc.)
writingStyle
Custom style instructions

Supported AI Providers

OpenAI

GPT-6 models (reasoning models). Strong at following clear instructions and structured output; temperature is managed by the model itself.

gpt-6.1-sol
gpt-6-astra
gpt-6-luna

Anthropic

Claude 5.5 models. Excellent for nuanced, long-form content and a consistent tone. Sonnet 5.5 is the default.

claude-sonnet-5-5
claude-opus-5-5
claude-haiku-5-5

fal.ai

FLUX image models for main and section images. FLUX.2 [pro] gives the best quality, FLUX.1 [schnell] the fastest drafts.

FLUX.2 [pro]
FLUX.2 [dev]
FLUX.1 [schnell]

ElevenLabs

Voice models for podcasts and voice-overs. Eleven v4 is the default: highly expressive, with audio tags such as [laughs] and natural multi-speaker dialogue. Romansh is not supported.

Eleven v4
Eleven v4 Turbo
Eleven v3
Eleven Multilingual v2
Eleven Flash v2.5