What inputs does the LLM Vendor Lock-In Cost need?

It takes 7 inputs: currentMonthlySpendUsd, promptComplexity, evalSuiteSize, retrainingEngineeringHours, downtimeDays, hourlyEngineeringCost, newVendorDiscountPercent. Outputs returned: totalSwitchingCost, monthsOfSpendEquivalent, paybackMonths.

Can I verify the LLM Vendor Lock-In Cost with a worked example?

Yes. With $8000 spend, complexity 6, 80 evals, 40 retrain hr, 2 downtime days, $150/hr, 30% discount. the tool returns totalEngineeringHours 104, switchingCost ≈ $16,133, paybackMonths ≈ 6.72.

Where does the LLM Vendor Lock-In Cost get its benchmark data?

Reference data is sourced from: BLS Occupational Employment and Wage Statistics — software developers (as of 2024-05); Stack Overflow Developer Survey (as of 2024).

What can the LLM Vendor Lock-In Cost not tell me?

Known limitations: Effort multipliers are heuristic; structured outputs, tool-call surface, and parser custom-code change real effort 2–10×. Doesn't model the cost of running both providers in parallel during cutover (a common pattern).

Methodology: LLM Vendor Lock-In Cost

1. Scope

Estimates engineering hours, downtime opportunity cost, and payback months for migrating between LLM providers. Heuristic effort multipliers, not measured velocity.

2. Inputs and outputs

Inputs

currentMonthlySpendUsd number ($)
promptComplexity number (1–10)
evalSuiteSize number
retrainingEngineeringHours number
downtimeDays number
hourlyEngineeringCost number ($)
newVendorDiscountPercent number (0–100)

Outputs

totalSwitchingCost

Engineering $ + downtime opportunity cost.
monthsOfSpendEquivalent

Switching cost / current monthly spend.
paybackMonths

Switching cost / monthly savings at new vendor discount.

Engine source: src/lib/llm-vendor-lock-in-cost/engine.ts

3. Formula / scoring logic

prompt_hr = complexity × 4
eval_hr = evals × 0.5
total_hr = prompt_hr + eval_hr + retraining_hr
downtime_$ = days × monthly_spend / 30
switching = total_hr × hourly + downtime_$

4. Assumptions

4 engineering hours per complexity-point of prompt rewriting.
0.5 hours per eval test for migration / recalibration.
Downtime opportunity cost approximated by daily share of monthly LLM spend — a placeholder when actual revenue impact isn't known.

5. Data sources

BLS Occupational Employment and Wage Statistics — software developers as of 2024-05
Stack Overflow Developer Survey as of 2024

6. Known limitations

Effort multipliers are heuristic; structured outputs, tool-call surface, and parser custom-code change real effort 2–10×.
Doesn't model the cost of running both providers in parallel during cutover (a common pattern).

7. Reproducibility

Input
$8000 spend, complexity 6, 80 evals, 40 retrain hr, 2 downtime days, $150/hr, 30% discount.

Expected output
totalEngineeringHours 104, switchingCost ≈ $16,133, paybackMonths ≈ 6.72.

8. Change log

2026-05-08 methodology first published.

Worked example

Run live against the same engine this site ships (/engines/llm-vendor-lock-in-cost.js). The inputs and outputs below are recomputed on every build and independently re-verified in CI — they are never hand-authored.

Input

tool: llm_vendor_lock_in_cost
current_monthly_spend_usd: 8000
prompt_complexity: 6
eval_suite_size: 80
retraining_engineering_hours: 40
downtime_days: 2
hourly_engineering_cost: 150
new_vendor_discount_percent: 30

Output

promptRewriteHours: 24
evalRewriteHours: 40
totalEngineeringHours: 104
engineeringDollarCost: 15600
downtimeOpportunityCost: 533.33
totalSwitchingCost: 16133.33
monthsOfSpendEquivalent: 2.02
monthlySavingsAtDiscount: 2400
paybackMonths: 6.72

Frequently asked questions

What does the LLM Vendor Lock-In Cost calculate?: Estimates engineering hours, downtime opportunity cost, and payback months for migrating between LLM providers. Heuristic effort multipliers, not measured velocity.
What inputs does the LLM Vendor Lock-In Cost need?: It takes 7 inputs: currentMonthlySpendUsd, promptComplexity, evalSuiteSize, retrainingEngineeringHours, downtimeDays, hourlyEngineeringCost, newVendorDiscountPercent. Outputs returned: totalSwitchingCost, monthsOfSpendEquivalent, paybackMonths.
What formula does the LLM Vendor Lock-In Cost use?: The exact computation is: prompt_hr = complexity × 4; eval_hr = evals × 0.5; total_hr = prompt_hr + eval_hr + retraining_hr; downtime_$ = days × monthly_spend / 30; switching = total_hr × hourly + downtime_$
Can I verify the LLM Vendor Lock-In Cost with a worked example?: Yes. With $8000 spend, complexity 6, 80 evals, 40 retrain hr, 2 downtime days, $150/hr, 30% discount. the tool returns totalEngineeringHours 104, switchingCost ≈ $16,133, paybackMonths ≈ 6.72.
Where does the LLM Vendor Lock-In Cost get its benchmark data?: Reference data is sourced from: BLS Occupational Employment and Wage Statistics — software developers (as of 2024-05); Stack Overflow Developer Survey (as of 2024).
What can the LLM Vendor Lock-In Cost not tell me?: Known limitations: Effort multipliers are heuristic; structured outputs, tool-call surface, and parser custom-code change real effort 2–10×. Doesn't model the cost of running both providers in parallel during cutover (a common pattern).