AI Cost & Savings Optimizer
Estimate your AI spend and get a ranked list of ways to cut it. Bring a usage CSV, scan a website, or paste your provider usage — all three feed the same breakdown and savings engine. Nothing you upload leaves your browser.
Export from your provider’s usage dashboard. Columns we read: model, input/output tokens, requests, cost, date, feature, user. Parsed in your browser — nothing is uploaded.
Frequently asked questions
How accurate is the AI cost estimate?
When you provide real usage (a CSV or your provider's usage API response), the model-level cost is computed directly from token counts and a maintained rate card, so it is close to your actual bill minus any custom discounts. The website mode is assumption-based: it can only see client-side AI usage, not server-side API calls, so it gives a ballpark from detected features and the volume you enter. Every number shows the rate and assumptions it used.
Is my usage data uploaded anywhere?
No. CSV parsing and the Connect-mode JSON both run entirely in your browser — nothing is uploaded or stored. The only server request is the optional website scan, which fetches the target site you name, not any of your data. Your provider API key, in Connect mode, never leaves your machine: you run the usage call yourself and paste the result.
What savings does it suggest?
Right-sizing frontier models to a cheaper sibling for simple tasks, prompt caching for repeated system context, moving non-realtime jobs to the Batch API (about half price), trimming output tokens on verbose responses, and provider arbitrage on frontier spend. Each recommendation shows a dollar-per-month impact, an effort level and the assumption behind it, sorted by impact.
Which providers and models does it cover?
The built-in rate card covers OpenAI (GPT-4o/4.1/o-series/3.5), Anthropic (Claude Opus/Sonnet/Haiku), Google Gemini, Meta Llama (hosted), and Mistral, including cached-input and batch tiers. Unknown models are flagged and excluded from computed cost. Rates are editable estimates — verify against your provider's current pricing.
Can it read my ChatGPT or Claude usage directly?
For the API platforms and Team/Enterprise admin accounts, yes — OpenAI and Anthropic expose usage/cost APIs whose JSON you can paste into Connect mode. Consumer ChatGPT and personal Claude plans have no usage API, so for those you would use the CSV export where available.
Rates are editable estimates (version 2026); verify against your provider’s current pricing. Savings are directional. Not financial advice.
Loading…
Sample report · illustrative data
Sample result: AI cost optimizer report
Example input: a pasted usage export (model, input tokens, output tokens, requests) from an invented team, analysed with the default assumptions of 40% right-sizable, 25% batchable and 30% cacheable input.
What the report shows
- Estimated spend
- Per month, scaled up from the period in your export
- Potential savings
- Up to a total per month, and as a percentage of spend, if every recommendation applies
- Where the money goes
- Cost and share of spend per model, with any model missing from the rate card listed as unpriced
Ranked savings (the recommendation types, each with its effort and assumption)
| Recommendation | Effort | Assumption it states |
|---|---|---|
| Right-size part of a frontier model to a cheaper sibling | Medium | The share of frontier-model spend that could use a cheaper model (default 40%) |
| Move non-realtime jobs to the Batch API | Low | The share of spend that is batchable (default 25%) at a 50% discount |
| Cache repeated system prompts and context | Low | The share of input tokens that is repeated context (default 30%) |
| Trim output tokens on verbose responses | Low | A 15% reduction in output tokens, shown when usage is output-heavy |
| Compare equivalent models across providers | High | A conservative 15% blended rate reduction on frontier spend, subject to quality parity |
How to read it
Savings are estimates from your numbers and the sliders you set; only you know your workload, so drag the assumptions and the savings update. Dollar figures depend on your export and the tool's rate card, so none are shown here.
Made-up example data. Your own run uses your inputs. Downloads of the full report need Pro.
More security tools
The tools people use most on PlayCISO.
Catch injected instructions hiding in content your AI agent should only read.
Score your readiness against the first international standard for AI security threats.
Critical CVEs across 21 major vendors, month by month — and what the 2026 surge means.
Build an ordered ruleset, replay labelled packets and catch shadowed and any-to-any rules.
Self-assess all six Functions, from Govern to Recover, and see your weakest one.
Search 172 threat actors across Microsoft, CrowdStrike, Mandiant and community names.