The Costiva Toolkit
Plan, price, and improve your AI — no account needed — answer a few questions and get a ready-to-use result. Every tool is free.
Plan your costs
Batch API Suitability Checker
Get a plain suitable / partially suitable / not suitable verdict for moving a workload to the Batch API.
Try it freeContext Window Budget Planner
See what fills up a model's context window each turn, and why chat sessions get expensive as history builds up.
Try it freeCost-Aware Routing Designer
Design a cheap-model-first cascade, price it honestly, and find the point where it stops being worth it.
Try it freeAI Cost Estimator Also listed under For Leaders
See what an AI-powered process could cost each month.
Try it freeModel Picker Also listed under For Leaders
Find out which AI model actually fits your task.
Try it freePrompt Caching Savings Analyzer
See what caching this exact prompt would save — and the point below which it costs more than it saves.
Try it freeRate Limit & Capacity Planner
See how close your peak traffic runs to your actual rate limits, and what to fix before you hit them.
Try it freeToken Counter
Know exactly how many tokens your text is before you send it.
Try it freeBuild your setup
Few-Shot Example Set Builder
Get the smallest example set that actually earns its keep — or find out you don't need one.
Try it freeMCP Config Wizard
Get a ready-to-use MCP config file for your AI client.
Try it freeSystem Prompt Generator
Turn a feature description into a production-ready system prompt.
Try it freeRAG Chunking & Retrieval Planner
Plan and price a RAG pipeline end to end — or find out full-context caching would be cheaper.
Try it freeReAct Prompt Converter
Convert a prompt into a ReAct loop, and see what it will cost before you run it.
Try it freeStructured Output Schema Builder
Get a JSON Schema, the code to validate it, and a repair prompt for when output drifts — all from a plain-language field list.
Try it freeSubagent & Slash Command Generator
Get a ready-to-save Claude Code subagent or slash command, scoped to least privilege.
Try it freeTool / Function Schema Generator
Turn a tool description into a valid schema the model calls correctly — and skips when it shouldn't.
Try it freeWorkflow to Agent Blueprint
Turn a manual process into a structured blueprint for automating it — honestly, not automatically.
Try it freeImprove what you have
Eval Runner
Prove a prompt or model change actually helped — with a pass rate, not a guess.
Try it freeGolden Dataset Manager
Turn scattered test cases into one versioned, reviewed asset your whole team trusts.
Try it freeGrounding / Hallucination Checker
Check whether an AI output actually says what its source document says.
Try it freeLLM-Judge Calibrator
Know whether your automated grading is actually trustworthy — before you rely on it.
Try it freePrompt Injection & Red-Team Tester
Find out whether your prompt can be manipulated — before a customer does.
Try it freePrompt Regression Diff
Catch a prompt regression before it ships — not after a customer does.
Try it freeAgent Trace Analyzer
Find exactly which steps in an agent session are burning money for no reason.
Try it freeSet your policies
Data Handling & Redaction Advisor
Know which data can go to which AI model or provider, and how to redact the rest.
Try it freeAI Incident Response Runbook Generator
Know exactly what to do in the first ten minutes of an AI incident — and have it printed before you need it.
Try it free