The free tier gives you everything to evaluate models and costs.
Pro adds personalised estimates, deeper analysis, and unlimited use.
Free
Everything you need to evaluate AI models and costs for your project.
Analysis
Cost breakdown
Providers & account
Free tier limits may change before launch
Pro
Deeper analysis, personalised estimates, and unlimited use for power users.
Deeper analysis
Cost tools
Customisation
Output & account
Pricing not final · Cancel anytime
| Feature | Free | Pro |
|---|---|---|
| Analysis | ||
| Project decomposition into task types | ✓ | ✓ |
| Recommended model per task with reasoning | 1 model | 3 tiers (Budget / Balanced / Premium) |
| Model capability badges (context window, multimodal, open-source, data residency) | — | ✓ |
| Cost breakdown | ||
| Per-1K token pricing | ✓ | ✓ |
| Savings vs GPT-4o everywhere | ✓ | ✓ |
| API vs subscription breakeven note | ✓ | ✓ |
| Cost estimate tiers | Medium scale only | Light / Medium / Heavy |
| Custom scale input (personalised estimate) | — | ✓ |
| Per-query & per-MAU cost framing | — | ✓ |
| Customisation | ||
| Providers considered | All 5 (default) | All 5 + restrict to specific vendors |
| Output & account | ||
| Projects | Up to 3 | Unlimited |
| Revisions per project | 3 | Unlimited |
Is the free tier actually free?
Yes — no credit card required. The free tier is not a trial; it's a permanent plan with a fixed feature set.
Can I cancel anytime?
Yes. You can cancel from the billing page inside the app at any time. You keep Pro access until the end of your current billing period.
What counts as a "project"?
A project is a named workspace for one application you're building. Each project stores your description and AI recommendation history. Free tier supports up to 3 projects with up to 3 revisions each.
What's the difference between a recommendation and an analysis?
You submit a project description and ModelSense returns a full recommendation — which models to use for each task, with cost estimates and reasoning. A revision is when you update your description and re-run the recommendation on the same project.
How accurate are the cost estimates?
Estimates are based on current published pricing from each provider. They're directionally accurate but depend on real usage patterns — token counts per request vary based on your prompts and data. Pro's custom scale input lets you enter your own usage numbers for a more precise estimate.
Can I request a feature or give feedback?
Yes — use the feedback button inside the app, or reach out directly. ModelSense is actively developed and feedback shapes the roadmap.