— Ingenious AI security check.
When you can’t leave a vendor, you can’t leave their security failures either — breaches, price hikes, and deprecated models become your problem with no exit. The calculator totals your switching costs: data gravity, retraining, integration rewrites. High lock-in means the vendor’s risk is your risk, permanently.
Check every AI tool you use, answer 4 depth questions, and get a 0–100 lock-in score plus an estimated switching cost in USD — with the math shown.
Check all that apply. Open-weights self-hosted counts against lock-in (it lowers the score).
1 = barely · 5 = deeply embedded. Tap one per question.
All figures are illustrative planning estimates, not quotes. Assumptions: data egress $0.12/GB (illustrative cloud egress rate), re-integration $1,800/engineering-week (illustrative), retraining $500/person (illustrative), early-exit fee 25% of remaining annual contract (illustrative), fine-tune retraining a flat $2,000–$10,000 by data depth (illustrative). Real costs vary with your data volume, architecture, contracts, and team.
TermsWatch monitors AI vendor terms and pricing changes — price hikes, egress-fee changes, deprecation notices, contract-term updates — so lock-in costs never move against you silently.
Our platforms are being upgraded and return November 1 — get early access now:
Get TermsWatch early access →The 0–100 score combines how many AI vendors you depend on with four depth dimensions: training/fine-tune data living with a vendor, number of API integrations, staff trained on vendor-specific tooling, and contract length. Higher scores mean a harder, more expensive switch. Open-weights self-hosted reduces the score.
From five illustrative components: data egress, re-integration engineering, staff retraining, contract early-exit fees, and fine-tune retraining. Every assumption is stated next to the result — the math is shown, not hidden.
No. All figures are illustrative planning estimates, not quotes. Real switching costs vary widely with your data volume, architecture, contracts, and team. Your mileage will vary.
TermsWatch monitors AI vendor terms and pricing changes — price hikes, egress-fee changes, deprecation notices, and contract-term updates — so lock-in costs don't move against you silently. It watches the fine print while you build.
When switching AI providers becomes expensive or risky because your workflows, data, or integrations depend on one vendor's proprietary features.
Score how deeply each tool is embedded — data gravity, custom integrations, retraining needs — then estimate migration effort. This calculator gives an illustrative 0–100 score with the math shown.
Prefer open standards and portable data formats, keep prompts and evals vendor-neutral, and document exit paths before you need them.
The 0–100 score combines how many AI vendors you depend on with four depth dimensions: training and fine-tune data living with a vendor, the number of API integrations, staff trained on vendor-specific tooling, and contract length. Higher scores mean a harder, more expensive switch.
From five illustrative components: data egress (GB moved out times an illustrative per-GB rate), re-integration engineering (estimated weeks times an illustrative weekly rate), retraining staff, contract early-exit fees, and fine-tune retraining cost. Every number is an illustrative estimate — not a quote — and the page shows the math.
No. All figures are illustrative planning estimates with assumptions stated in plain sight. Real switching costs vary widely with your data volume, architecture, contracts, and team. Your mileage will vary.
TermsWatch monitors AI vendor terms and pricing changes — price hikes, egress-fee changes, deprecation notices, and contract-term updates — so lock-in costs don't move against you silently. It watches the fine print while you build.