Real-world stories from developers who reduced their Claude Code costs using the strategies in this repo.
Case studies document the full journey of a cost optimization effort: what the project looked like before, what changes were made, and the measurable results after. Unlike tips (which describe a single technique) or benchmarks (which isolate a specific variable), case studies capture the combined effect of multiple optimizations applied to a real project over time.
Good case studies include:
- Concrete numbers -- actual token counts, dollar amounts, and percentage reductions
- Context -- project size, tech stack, team size, and how Claude Code was being used
- Specifics -- exactly which optimizations were applied and in what order
- Honest assessment -- what worked, what didn't, and any tradeoffs encountered
| # | Title | Stack | Savings | Date |
|---|---|---|---|---|
| -- | Coming soon | -- | -- | -- |
No case studies have been published yet. Be the first to submit one.
There are two ways to contribute:
Use the Case Study Submission issue template. Fill in the prompted fields with your data. The maintainers will review it and, if accepted, convert it into a published case study file.
- Copy the TEMPLATE.md file in this directory.
- Rename it to something descriptive (e.g.,
react-saas-app.mdormonorepo-migration.md). - Fill in all sections with your data.
- Open a PR targeting
main. See CONTRIBUTING.md for general PR guidelines.
- Specific numbers over vague claims. "$8.50/day down to $3.20/day" is better than "much cheaper."
- Before and after data. Include token counts, session costs, or monthly spend for both states.
- Reproducible details. Describe your setup clearly enough that others with similar projects can apply the same approach.
- Honest scope. A small but well-documented optimization is more useful than a vague claim of massive savings.
Every case study should follow the structure defined in TEMPLATE.md:
- Project Overview -- type, size, stack, team
- Before Optimization -- baseline costs and pain points
- What Changed -- specific optimizations applied
- After Optimization -- new metrics
- Results Summary -- savings percentage and implementation effort
- Lessons Learned -- what worked, what didn't, advice for others
Tables are preferred for before/after metric comparisons. See the template for the exact layout.