Time Machine Help

How It Works

• Pick two AI providers — one to Answer and one to Review. Choose from OpenAI, Anthropic, xAI, Perplexity (with Llama and Mistral coming soon).
• Enter your prompt — the Answer AI generates a response, then the Review AI critiques it with feedback and suggestions.
• Get two perspectives — see where providers agree and where they differ, giving you a more complete picture.
• Save snapshots — every result is saved so you can revisit and compare past responses over time.

Compare Mode

• Select 2 or more providers using the checkboxes to run the same prompt across all of them simultaneously.
• Results appear side by side so you can see how each provider handles the same question.
• Great for evaluating which AI is best for specific tasks — research, coding, writing, etc.

Best Practices

• Use Perplexity as Reviewer for fact-checking — it searches the web to verify claims.
• Use Claude as Reviewer for logic and reasoning checks on complex analysis.
• Use Compare Mode to benchmark providers before choosing one for a workflow.
• Save important snapshots with clear names — they become your reference library.

Troubleshooting

• Provider not responding? Check that the provider is configured in Settings with valid credentials.
• Review AI not critiquing? Make sure you've selected a different provider for Review than for Answer.
• Snapshots not saving? Ensure the server is running and has write access to the data directory.
• "Coming Soon" providers greyed out? Llama and Mistral are not yet available — they'll be enabled in a future update.