Reads the log
Uses the data your setup already produces.
Tune your engine in minutes
Start to finish in one workflow—not a throwaway starter map.
Baseline. VE, lambda, spark from physics.
Refinement. Logs: commanded vs actual lambda, airflow vs model—VE, trim, spark.
Tuned. Confirmed on the car.
Create an account and start a plan (DIY monthly includes a 7-day trial) to add vehicles and run reports.
CT
Analysis
Pull #12 · WOT hit
Ready
Vehicle profile
1969 Camaro
502cid · Holley Dominator · E85
- Goal
- Street / occasional strip
- Last log
- pull_20260325.csv
AI-assisted read
Likely issue
Lean spike post-shift
Confidence
78%
Recommended next checks
- Verify fuel pressure under load vs. demand
- Compare commanded vs. measured lambda around shift event
- Inspect transient fuel / accel enrichment tables
The full arc
We do not stop at “first start”—we work toward tuned.
Physics-first baseline when you need it—VE, lambda, spark—or start from CSV logs and refine commanded vs actual lambda and airflow vs the model into VE, trim, and spark you re-log to confirm. Goal: tuned, not just started.
-
1
Combo profile → baseline math
Structured setup grounds VE, lambda, spark, and startup strategy in your hardware—not a generic basemap thread.
-
2
Datalogs → corrections
Upload CSVs from your ECU toolchain; we align channels to what the car actually did and suggest VE, fuel trim, and spark moves with clear verification steps.
-
✓
Tuned
When the data and the seat-of-the-pants agree, you are done—not when the starter map merely lights the engine.
Create an account and choose a plan to save vehicles and run reports.
Assistance only — verify on the car and follow your ECU manufacturer’s guidance.
Overview
We are built around two kinds of truth: math that gets you in the ballpark, and data that finishes the job. Start from a deterministic baseline or jump straight to logs—either way the product pushes you toward the same outcome: a calibration that matches the real engine.
01
Deterministic baseline
VE from the engine model, lambda targets from fuel and load, and a conservative spark curve—so the first pull is grounded in physics, not vibes.
02
Adaptive refinement
Stack datalogs and close the gap: commanded vs actual lambda, airflow vs prediction—then VE, fuel trim, and spark optimization with clear next checks.
03
Tuned
Not another “step”—the outcome. A map state that keeps improving with each verified pass until the car and the data agree.
04
Built for DIY and pros
Same pipeline for your garage project or a shop that lives in CSV exports every week.
Not just another chatbot
Deterministic math. Not guesswork.
Generic AI chatbots generate text about tuning. Crazy Tuner calculates your tune from physics — ideal gas law, fuel stoichiometry, injector dynamics, and thermodynamics — then uses AI to interpret your build and explain what to check.
Generic AI chatbots
ChatGPT, Grok, Gemini, etc.
- Generates narrative text about tuning — no actual calculations behind the numbers
- No knowledge of your specific build — gives the same generic answer to every combo
- VE, lambda, and spark values are made up — statistically plausible text, not physics
- Cannot read your datalog or verify its own output against real sensor data
- Confidently wrong — will hallucinate injector sizes, AFR targets, and timing numbers
"Give me a VE table for my LS turbo on E85" → you get a wall of text with numbers that look right but were never calculated from anything real.
How we do it
Crazy Tuner
Deterministic math + AI
- Physics engine calculates VE, lambda, spark, fuel mass, and injector PW from thermodynamics
- Every cell is computed from your specific build — displacement, cam specs, injector flow, fuel type, boost target
- Ideal gas law for air charge, stoichiometric ratios per fuel, shaped VE surfaces — real math, not token prediction
- AI reads your datalog, understands your combo context, and explains what to check — with the math already done
- Output is validated — IDC limits, VE monotonicity, and safety thresholds are checked before you see it
Same LS turbo on E85 → you get a VE table, lambda target, and spark map calculated from your bore, stroke, cam duration, injector size, and boost target. Every cell traces back to a formula.
Physics-first
Engine math from your combo through maps, with safety checks — runs automatically in the product, not as a long manual checklist.
0 hallucinated
cells — every value is calculated, not predicted
Your build
not a generic LS/Coyote/2JZ template
The problem is not a lack of data. It is knowing what to do next.
Aftermarket tuning can turn into a cycle of log review, repeated revisions, and expensive trial and error.
DIY builders often know enough to work on their own car, but not enough to always know the next best move with confidence.
Shops and professional tuners often know exactly what they are doing, but still lose time to repetitive log review and manual diagnostics.
Crazy Tuner helps both groups move from baseline to refinement to a tune they can stand behind—without drowning in noise.
Without Crazy Tuner
- — stare at logs
- — guess likely causes
- — waste time on bad assumptions
- — pay for or perform unnecessary revisions
With Crazy Tuner
- ✓ lock in a physics-first baseline when you need it
- ✓ compare commanded vs actual and model vs measured airflow
- ✓ apply VE, trim, and spark corrections with reasoning
- ✓ iterate until the calibration is actually tuned
A tuning copilot, not a blind auto-tuner
Crazy Tuner pairs a physics engine for the baseline with an AI copilot for the refinement loop—so you always know whether the next move is about the model, the sensors, or the tables.
- It does not replace your ECU.
- It does not recklessly rewrite your tune.
- It helps you run the full arc: safe starting maps, structured log review, and map-state updates until you are tuned—not just “started.”
Understands the build
Considers engine, fuel system, drivetrain, and intended use.
Guides the next move
Shows what likely happened, why it matters, and what to check next.
Workflow
How you get from baseline to tuned
Two engineering phases—math first, then your datalogs—with tuned as the finish line, not a mystery step three weeks later.
-
Math and physics
1Deterministic baseline
You: Save your combo profile (engine, ECU, fuel, goals). New build? Generate a base-tune guide. Already running a map? You can skip straight to logs—same vehicle context either way.
Crazy Tuner: We derive a VE table from the engine model, lambda targets from fuel type and load, and a conservative spark curve—plus startup strategy and a first-start checklist—so the car has a defensible starting point before you chase edge cases.
-
Data from the car
2Adaptive refinement
You: Export CSV datalogs after each session. Read the report, make careful table changes in your ECU software, and re-log until the errors you care about shrink.
Crazy Tuner: We line up commanded lambda vs actual, airflow vs what the model predicted, and surface where VE, fuel trims, or spark should move—with confidence, reasoning, and map-state updates so every pass builds on the last.
-
Not “step 3”—the outcome
Tuned
When the commanded and measured stories agree, drivability feels right, and you trust the margins—that is tuned. Crazy Tuner is built to keep you in that loop until you get there, not to drop you off after a starter map.
Built for two types of users
DIY
DIY Builders
For capable enthusiasts who want guidance, diagnostics clarity, and smarter recommendation support without paying for endless revisions.
See pricingProfessional
Professional tuners
For professionals who want faster log triage, less repetitive diagnostics, and a more efficient tuning workflow.
See pricingBuilt to help you tune smarter, not recklessly
Crazy Tuner is designed to improve decision-making, not blindly make risky changes.
- Confidence-based recommendations
- Hardware-verification prompts
- Clear next-step guidance
- Escalation when expert review is needed
- Human judgment stays in the loop
More performance applications are coming
Crazy Tuner is starting with a focused launch and expanding into broader performance workflows over time. Roadmap priorities follow customer demand for additional ECU export formats, shop collaboration features, and tighter integration with how professional tuners already work in the garage and on the dyno.
Pricing built for how people actually use the product
Two simple subscriptions: DIY and Professional. Each includes a monthly report allowance (base tune or log analysis). If you hit the cap before the month resets, add prepaid reports from billing—Professional includes a higher allowance for busy calibrators.
DIY
$25/mo or save 15% yearly. DIY monthly includes a 7-day trial before you are billed. Same Holley, MoTeC, Haltech, FuelTech, and HP Tuners CSV workflow—throttled by fair monthly usage, not confusing pack ladders.
Professional
$200/mo or save 15% yearly with a much higher included monthly allowance. Built for throughput, customer handoffs, and teams that live in logs every week.
FAQ preview
- What is Crazy Tuner?
- A copilot for the whole arc: physics-first baseline maps (VE, lambda, spark), then log-driven refinement until you are tuned—not just a starter file.
- Do I need a base tune if I already have a calibration?
- No. Upload CSV datalogs and refine what is on the car. The same profile ties together baseline generation and closed-loop corrections.
- Is it for DIY or professionals?
- Both — the same baseline-to-tuned story for builders and throughput-focused workflows for shops.
- Does it replace a tuner?
- No. It accelerates clarity and diagnostics; expertise stays in the loop.
- Is it automatic tuning?
- No. Recommendations are explained and confidence-scored — not blind table rewrites.
Baseline, refinement, tuned.
Builders and shops use the same pipeline: defensible starting maps when you need them, structured log review when the car is lying to the model, and a map state that improves until you are done—not until you “got it running.”
Sign up with email, add a vehicle profile, then run a guided base tune, jump straight to CSV log analysis, or alternate between both as the combo evolves. Questions on usage, safety, or billing are on the FAQ and pricing pages.