AI Cost Optimization Playbook
Cut your LLM costs by 60-90%. Without sacrificing quality.
5
Skills
5
Layers
60-90%
Typical savings
What you'll be able to do
Concrete outcomes, not vague promises.
Your LLM bill is growing faster than usage
Last month $2k, this month $8k, next month projected $30k. You don't know where the money is going. Finance is asking questions.
You're using the most expensive model for everything
Sonnet for every call, including simple classification. Haiku would work for 80% of your calls at 1/30 the cost.
You're building for scale
1k users today, 100k next quarter. At current cost, 100k users = $50k/month. You need a cost-aware architecture from day 1.
Built for
If any of these are you, this is for you.
Why this playbook works
Three things that make this different from a wall of prompts.
The 5-layer funnel
Cache first (50-90% reduction). Token optimization second (20-50%). Model selection third (50-80%). Batch API fourth (50% off). Fine-tune last. Apply in order.
Most teams find 80% waste
The first audit almost always reveals 80% waste. Unnecessary calls, oversized prompts, expensive models for simple tasks. The pack teaches you to find it.
Quality preservation
Cost optimization doesn't mean quality loss. The pack teaches you to measure quality (via evals) and optimize cost with the constraint that quality doesn't drop.
Quick start
Drop the package into your project and start using the commands.
What's in the box
- 5 skillsfunnel + 4 layer skills
- 5 tool adapterspi-agent, Claude Code, Cursor, Cline, generic
- 2 worked examplestoken reduction + budget alerts
- Cost math per layerknow the savings
- Vendor-agnosticworks with any LLM provider
Everything in the pack
What you get the moment your payment clears.
Adapters included
These skills were authored for pi-agent, but we include drop-in adapters for the major AI coding tools.
| Tool | Adapter location | What you get |
|---|---|---|
| pi-agent | adapters/pi-agent/ | SKILL.md drop-in |
| Claude Code | adapters/claude-code/ | .claude/commands/*.md + CLAUDE.md |
| Cursor | adapters/cursor/ | .cursorrules rule sets |
| Cline / Continue | adapters/cline/ | System-prompt rules |
| Aider / generic | adapters/generic/ | INSTRUCTIONS.md snippets |
Same playbook, different packaging. The adapters translate the methodology into the format each tool expects. Install once, swap agents without re-learning the method.
Full documentation
Everything in the package, every prop, every pattern.
AI Cost Optimization Playbook
Cut your LLM costs by 60-90%. Without sacrificing quality.
Built and battle-tested. 5 skills covering token economics, model selection, caching strategies, and FinOps for LLMs. Adapters for pi-agent, Claude Code, Cursor, Cline.
What's Inside
| Skill | Purpose |
|---|---|
| cost-optimization | Master orchestrator — the cost discipline, where money is hiding, the optimization funnel |
| token-economics | Token math: input vs output, prompt caching, context window economics, batch pricing |
| model-selection | Picking the right model, cost/quality tradeoffs, distillation, fine-tuning economics |
| caching-strategies | Prompt caching, semantic caching, response caching, when each is right |
| finops-for-llms | Cost attribution, budgets, alerts, FinOps discipline applied to LLM workloads |
Quick Start
# 1. Extract
tar -xzf cost-optimization-0.1.0.tgz
cd cost-optimization-0.1.0
# 2. Pick your agent host
ls adapters/
# pi-agent (recommended)
mkdir -p ~/.pi/agent/skills
cp -r skills/* ~/.pi/agent/skills/
# Claude Code
mkdir -p .claude/commands
cp -r adapters/claude-code/commands/* .claude/commands/
The Core Idea
The single sentence that ties this whole pack together:
The cheapest LLM call is the one you don't make. The second cheapest is the one that's 100x smaller.
The pack teaches the cost optimization funnel: stop unnecessary calls (caching, deduplication) → reduce call size (token optimization) → use cheaper models (model selection) → batch where possible (batch API) → only then fine-tune. Every layer saves 30-60% when applied.
The 5 Skills in Detail
cost-optimization — The orchestrator
- The optimization funnel (cache → token → model → batch → fine-tune)
- Where money hides (most teams find 80% waste on first audit)
- The "is it worth it?" check
token-economics — Stage 1
- Input vs output cost (output is 4-5x more expensive)
- Prompt caching (Anthropic) — 90% cost reduction on cached reads
- Context window economics
- Batch API pricing (50% off)
model-selection — Stage 2
- Cost/quality tradeoff matrix
- When to use Haiku vs Sonnet vs Opus
- Distillation (fine-tune small model from large)
- Fine-tuning economics (when it's worth it)
caching-strategies — Stage 0 (the biggest win)
- Prompt caching (provider-level)
- Semantic caching (caching similar queries)
- Response caching (caching identical queries)
- When each is right (and when it's wrong)
finops-for-llms — The discipline
- Cost attribution (per user, per feature, per team)
- Budgets and alerts
- Showback vs chargeback
- Vendor management (rate negotiations, volume discounts)
Adapters Included
Works in every major AI coding tool:
| Tool | Adapter location | What you get |
|---|---|---|
| pi-agent | adapters/pi-agent/ | SKILL.md drop-in |
| Claude Code | adapters/claude-code/ | .claude/commands/*.md + CLAUDE.md fragments |
| Cursor | adapters/cursor/ | .cursorrules rule sets |
| Cline / Continue | adapters/cline/ | System-prompt rules |
| Aider / generic | adapters/generic/ | INSTRUCTIONS.md snippets |
Updates
Free for life. Re-download for new model pricing, new caching techniques, new cost patterns.
What This Isn't
- Not a cost-monitoring tool — methodology, not infrastructure
- Not specific to one vendor — applies to OpenAI, Anthropic, Google, open-weight
- Not "use a smaller model" — that's only one of 5 layers
What's in the download
cost-optimization-0.1.0/
├── README.md
├── OVERVIEW.md
├── METHODOLOGY.md
├── INSTALL.md
├── LICENSE.md
├── package.json
├── skills/ ← 5 SKILL.md files
│ ├── cost-optimization/
│ ├── token-economics/
│ ├── model-selection/
│ ├── caching-strategies/
│ └── finops-for-llms/
├── adapters/
├── examples/ ← 2 worked examples
└── docs/
Common questions
Things people ask before buying.
Does this work with my LLM provider?
Yes. The methodology is vendor-agnostic. Apply the same funnel to OpenAI, Anthropic, Google, open-weight, or any combination. The specific tactics differ (Anthropic has prompt caching, OpenAI has batch API); the layers are the same.
Will cutting costs hurt quality?
Not if you measure quality via evals. The pack teaches you to optimize cost with the constraint that eval scores don't drop. Most teams find 60-90% cost reduction with no measurable quality change.
How long does the first audit take?
A focused first-pass audit (find the biggest waste) takes a day. Applying the fixes takes 1-2 weeks. The pay-back is usually within the first month of cost savings.
£35 — lifetime
One purchase. Yours forever.
Pay once via Stripe. Get the complete package instantly. Free updates for the same major version. Commercial license included.
Secure payment via Stripe · Instant download · 30-day money-back if it doesn't save you time