Cut your AI dev tools bill by 60% — without cutting headcount.
Your engineers' coding agents burn most of their tokens re-reading the same codebase every session. Lettuce gives the agent a map so it stops paying for rediscovery. Same engineers, same agents, same output — a much smaller invoice.
You're paying for the same lookup, over and over.
Coding agents have no memory between sessions. Every prompt, the agent starts from zero — greps the repo, opens random files, re-reads modules it read yesterday. Most of the input tokens on your invoice go to context the agent could have looked up once.
A rediscovery tax on every session
Roughly 60% of input tokens on a typical task go to the agent finding its way around — measured on real GitHub issues, not modeled.
The bill scales with engineers, not output
Hire more developers, run more sessions, your AI invoice grows linearly. The rediscovery tax compounds with every seat.
No visibility into where it goes
The Anthropic / OpenAI invoice is one line item. You can't tell which engineer, which repo, or which task drove it.
Drop in last month's bill.
We apply the same input-token reduction we measured on real GitHub issues to your number. No need to guess at sessions, developers, or models — the ratio holds regardless.
Whatever your team is paying Anthropic, OpenAI, Cursor, etc. for agent usage right now. Pull it from last month's invoice.
Savings ratio comes from the published benchmark (Claude Code solving real GitHub issues with and without Lettuce — same methodology, see below). It's applied flat to whatever you're spending today, because Lettuce cuts input tokens regardless of the model on the other end.
How we compute “tokens saved”+
We don't ask you for your “before” spend and we don't make it up. Every Lettuce tool call writes two numbers to the same row in our database:
- served — measured. The exact byte size of the response Lettuce returned to your agent, converted to tokens.
- baseline — modeled. What the agent would have read from source to answer the same question without us.
tokens saved is max(baseline − served, 0) summed across every call. The clamp matters: if a call ever returns more than its baseline, the row contributes zero — never a negative “saving.”
How baseline is computed per tool:
read_snippet— near-exact. The baseline is the full file the slice came from; the saving is everything we trimmed off.- navigation tools (
find_symbol,explain_symbol,callers) — conservative. The baseline is the combined source span of the symbols the call returned: the lines the agent would have had to read to locate them. This ignores the surrounding file the agent would also have pulled in, so we under-count. The real saving is higher.
The calculator on this page uses the same numbers, averaged across 99 real GitHub issues solved by Claude Code with and without Lettuce (14,969,244 baseline tokens vs. 5,934,465 served, a 60% reduction). See the full run.
Predictable spend, line-item visibility, no engineering risk.
Lettuce is an MCP server. Your engineers paste a URL + key into their coding agent and continue working. Nothing in their workflow changes. The savings show up on the next invoice.
A smaller invoice from the providers you already pay
Anthropic, OpenAI, whoever powers your agents — fewer input tokens means a smaller bill. Same models, same caps, ~60% less spend.
Per-team budgets and caps
Set spend ceilings per team or repo. Stop a runaway agent before it costs you four figures over a weekend.
Usage reports you can actually read
Today we show calls by tool and tokens saved. Spend split by repo / team / agent is not built yet.
Zero engineering risk
If Lettuce disappears tomorrow, agents fall back to grepping the repo. No lock-in. No migration. Engineering can stop using it on a Tuesday.
Built for how you actually buy software.
- Annual contracts, NET-30/60, MSAs accepted
- SSO (SAML / OIDC) + SCIM on Enterprise
- Self-host in your VPC — source never leaves
- Per-team / per-repo spend caps and usage reports
- Audit logs of every agent call
- DPA, security questionnaire, vendor onboarding
Want the spreadsheet version?
We'll walk through your last 90 days of agent invoices, project the savings on your actual numbers, and send a one-page ROI memo you can forward to finance.