Configuration¶

OpenAI Model Configuration¶

FlowAgent uses OpenAI's language models for workflow generation and analysis. Different operations have different model requirements:

Workflow Generation (gpt-3.5-turbo or better)¶

Basic workflow creation and execution can use gpt-3.5-turbo
Set in your .env file:
```
OPENAI_MODEL=gpt-3.5-turbo
```

Report Generation (gpt-4-turbo-preview recommended)¶

For comprehensive analysis and insights, use gpt-4-turbo-preview
This model provides better reasoning and analysis capabilities
Set in your .env file:
```
OPENAI_MODEL=gpt-4-turbo-preview
```

Example configurations:¶

For workflow execution:

# Set model in .env
OPENAI_MODEL=gpt-3.5-turbo

# Run workflow
flowagent "Analyze RNA-seq data in my fastq.gz files using Kallisto..."

For report generation:

# Set model in .env
OPENAI_MODEL=gpt-4-turbo-preview

# Generate comprehensive analysis
flowagent "analyze workflow results" --analysis-dir=results

You can also set the model temporarily using environment variables:

# For one-time report generation with gpt-4-turbo-preview
OPENAI_MODEL=gpt-4-turbo-preview flowagent "analyze workflow results" --analysis-dir=results