Know the real cost of a price change before reacting
Providers raise and cut LLM prices regularly. This calculator turns a percentage announcement into the figure that actually matters: the dollar change to your monthly and annual bill — and how much a cheaper hedge model would save if you switched.
Why providers change prices and what to watch for
LLM pricing has moved significantly as the market matures — both upward on premium models as capabilities improve, and downward on older or smaller models as competition increases and inference costs fall. A typical pattern: a new flagship model launches at a high rate, then cuts are announced 6–12 months later as the next generation arrives and the previous model becomes the mid-tier offering.
What matters to you is not the percentage itself but what it means in dollars on your actual usage. A 30% price increase on a model you spend $50/month on is a $15/month problem — worth noting, not worth a migration project. The same 30% increase on a $4,000/month bill is $1,200/month or $14,400/year — very likely worth a sprint to evaluate alternatives or optimize usage.
How the impact is computed
A price change scales your bill directly:
new_monthly = current_monthly x (1 + pct_change / 100)
annual_delta = (new_monthly - current_monthly) x 12
For example, a 25% increase on a $1,200/month bill adds $300 a month and $3,600 a year. A 20% price cut on the same bill saves $240 a month and $2,880 a year — potentially freeing budget to expand usage or upgrade quality without spending more overall.
The hedge model field
The hedge field lets you compare your post-change cost against what the same workload would cost on an alternative model. This is most useful for answering: “Is the saving from switching worth the engineering effort?” If switching saves $8,000/year and the migration takes two weeks of a senior engineer’s time, that is likely worthwhile. If it saves $400/year, it probably is not.
Keep a cheaper alternative model always validated against your core tasks so that when a price change arrives, a migration is a config edit rather than a research and testing project.
Tips for responding to price changes
- React to the annual number, not the monthly one. Small monthly deltas compound into budgets worth acting on.
- Keep a hedge model qualified. Having a cheaper alternative already tested means a price hike is a config change, not a project.
- Bank or reinvest cuts deliberately. A price drop frees budget — decide whether to save it, grow usage, or upgrade quality, rather than letting it drift.
- Check if the pricing change affects input, output, or both. Providers sometimes change input and output rates independently — re-run your actual input/output ratio through the pricing calculator to get the real impact.