6.6L Duramax Horsepower: LML and L5P Source Guide

Direct answer: verified output depends on the Duramax generation

The reviewed GM evidence supports 397 hp at 3,000 rpm for the 2016 Silverado HD LML application and 470 hp at 2,800 rpm for GM's latest L5P Silverado HD and Sierra HD test application. State the engine code, vehicle context, source year and peak speed whenever using either number.

Do not answer a family-wide question by silently choosing the newest headline. A 2002, 2006, 2009, 2016, 2019 and current HD diesel may all carry a 6.6L Duramax description while belonging to materially different generations and truck systems. Unsupported rows stay blank.

Begin with the engine code, not displacement alone

Record the private VIN, model year, build date, Chevrolet or GMC vehicle, pickup or chassis-cab identity, engine code, transmission and emissions configuration. The shared 6.6-litre displacement is not enough to choose a horsepower row.

Use build records, the engine label and exact-year GM literature together. A registration year can miss a production change, while an exterior badge cannot prove that the installed engine, controller or calibration still matches the original build.

The 2016 LML figure is an exact application anchor

GM's official 2016 Silverado HD specifications identify the Duramax 6.6L turbo-diesel V8 as an LML and publish 397 hp at 3,000 rpm. That row is useful because the engine code, model family and model year travel with the value.

It does not automatically establish the same rating for every LML year, GMC application, Canadian-market truck, chassis cab or replacement engine. Preserve the row as reviewed and retrieve exact literature before filling a nearby application.

The latest L5P figure is not a universal historical rating

GM's current L5P technical reference publishes 470 hp at 2,800 rpm. Its testing footnote scopes the result to Chevrolet Silverado Heavy Duty and GMC Sierra Heavy Duty. Keep that application statement beside the headline.

The same page compares the latest engine with 2023-and-older L5P engines. That wording is direct evidence that L5P itself is not one timeless calibration. Do not backfill 470 hp into 2017, 2019, 2023 or another unreviewed L5P row.

The 73-hp gap is labelled arithmetic, not a trend

Subtracting the reviewed 397-hp LML anchor from the reviewed 470-hp latest-L5P anchor gives 73 hp. Dividing 73 by 397 gives about 18.4 percent. These are transparent editorial calculations between two source endpoints.

They do not establish an annual improvement rate, isolate the effect of one component or describe the missing generations between them. Do not interpolate unsupported horsepower values or claim every newer truck produces the full endpoint difference.

Peak speed belongs to the exact horsepower row

The reviewed LML row places 397 hp at 3,000 rpm; the current L5P reference places 470 hp at 2,800 rpm. Keep each rpm attached to its horsepower rating. Moving either speed to another generation creates a specification that GM did not publish in the reviewed evidence.

A peak point does not reveal the full power curve, shift strategy or output at another engine speed. Do not draw a plateau, estimate area under the curve or infer acceleration from two isolated peaks.

Torque context must not replace the horsepower answer

GM's current L5P reference also publishes 975 lb-ft at 1,600 rpm and says 90 percent of peak torque is available from 1,550 through 2,850 rpm. Those details help identify the current application but answer a separate torque question.

Do not attach 1,600 rpm to the 470-hp peak, convert the torque-band statement into another certified horsepower value or use the latest torque figure for the 2016 LML row. The dedicated torque guide retains that intent.

Do not manufacture a complete Duramax generation table

A useful reference shows what has been verified and what remains unresolved. The present primary evidence supports one 2016 LML application and GM's latest L5P test application. It does not yet populate LB7, LLY, LBZ, LMM, every LML year or earlier L5P calibrations.

Leaving a cell blank is more accurate than copying an enthusiast chart or assuming neighbouring years match. Add a generation only after reviewing a manufacturer specification, original brochure or other primary application-specific record.

Current L5P architecture stays L5P-scoped

GM's L5P reference lists 6,604 cc displacement, a 16.0:1 compression ratio, 103-by-99-mm bore and stroke, and high-pressure common-rail direct injection for the latest engine it describes. Those fields can identify that reviewed configuration.

They must not be copied into earlier Duramax rows merely because every engine shares the 6.6L family label. Architecture, materials, turbocharging, fuel system and emissions equipment need generation-specific sources and belong in the consolidated specifications workflow.

Factory engine horsepower is not wheel horsepower

The GM ratings are engine-output specifications, not chassis-dynamometer results. Wheel output depends on the transmission, converter state, transfer case, axle ratio, tires, losses, correction method, test gear, ambient conditions and current mechanical state.

Do not apply a generic driveline-loss percentage to invent a wheel figure. If a candidate is measured, retain the equipment, correction standard, gear, converter state, tire details, calibration and conditions, and label the result as measured rather than factory engine horsepower.

Horsepower does not set towing capacity

A Silverado or Sierra trailering rating comes from the complete truck configuration: series, cab, bed or body, drivetrain, axle ratio, tires, hitch type, equipment and certification limits. Engine horsepower is only one system input.

Use the exact candidate's labels and matching Canadian or U.S. configuration table as appropriate. Account for passengers, cargo, accessories, hitch and trailer loads, and let the lowest applicable rating control. Do not use 470 hp to raise a payload, axle, tire, hitch, GCWR or trailer limit.

Horsepower alone cannot predict acceleration

Acceleration depends on the full power and torque curves, gearing, shift control, vehicle mass, tires, traction, aerodynamics, grade, temperature and condition. A larger peak number does not define launch, passing or grade performance for unmatched trucks.

Compare performance only under a defined protocol with matched configurations. Never infer elapsed time, speed, safe merge distance or loaded performance from a brochure peak, and never turn a subjective drive into a repeatable specification.

Modifications create a separate output record

A tune, turbocharger, fuel-system, intake, exhaust, emissions, transmission or tire change can alter output and evidence requirements. Claimed gains cannot be added to a factory number, and a parts list does not prove the installed calibration or measured result.

Record exact hardware, software version, installer, date and supporting documentation. Preserve the appropriate factory row as historical context. This guide does not approve a modification, determine legal compliance, predict durability or certify a current candidate's output.

A derate or warning requires diagnosis, not arithmetic

Reduced-power messages, warning lamps, smoke, fuel-pressure concerns, boost faults, temperature problems or transmission behaviour can affect how a truck performs. Subtracting a guessed percentage from 397 or 470 hp does not diagnose any of them.

Retain codes, freeze-frame data, operating conditions, service history and test results. Follow exact GM service information or qualified diagnosis. Do not promise that one service or part will restore a headline rating without a confirmed cause and comparable measurement.

Horsepower cannot establish reliability or lifespan

An output rating contains no failure-rate denominator, exposure, severity or outcome definition. It cannot rank generations, name a best year, identify years to avoid or predict remaining life. Those questions require separate evidence and candidate inspection.

Keep problems, reliability, best-years and years-to-avoid pages independent. A truck can retain its factory calibration while having unresolved condition, and a measured shortfall may have many causes that the horsepower number cannot distinguish.

Canadian VIN applicability requires Canadian evidence

Use Chevrolet Canada's owner-literature and recall pathways plus the Government of Canada recall database with the exact private VIN. Retain the dated result and completed-repair record. A U.S. specification row is not automatically a Canadian VIN statement.

Those lookup systems do not certify installed calibration, horsepower, condition, reliability, inventory, seller, dealer count, price, financing or offer. If market or chassis calibration remains unclear, keep the field unresolved and obtain an application-specific GM record.

Build a reproducible horsepower record

A reproducible comparison lets another reviewer reach the same conclusion from the same truck identity and source. Keep factory, calculated and measured values separate, retain publication dates and record every unresolved field instead of hiding it behind one family headline.

When two sources conflict, preserve both contexts and investigate the application difference. Do not average ratings, select the more attractive value or use a current technical page to overwrite an original model-year specification.

Use horsepower evidence during a used-truck inspection

Compare the build identity, engine code, calibration evidence, emissions equipment, transmission, modifications, warnings, service history and diagnostic state with the correct source row. Photograph identifiers and request invoices for engine or controller changes.

Do not run an unsafe or unresolved truck merely to test performance. Obtain qualified inspection and diagnosis. Price only defined maintenance and diagnosed repairs; do not invent a market discount, reliability score or remaining-life penalty from a horsepower claim.

Separate the supporting Duramax workflows

This page owns generation- and application-scoped horsepower. The specifications page consolidates supported architecture and service fields; the torque page owns torque output; maintenance and fluid pages own exact service instructions; problems and reliability pages own their separate evidence.

Keeping those intents separate prevents a current output headline from swallowing a service, diagnosis, towing, used-buying or year-selection question. Link between the pages, but do not duplicate or canonicalize one distinct intent into another.

6.6L Duramax horsepower checklist

Complete this sequence before applying a horsepower figure to an engine or truck.

• Identify the private VIN, model year, build date, GM vehicle and completed body.
• Verify the engine code and installed engine rather than relying on 6.6L alone.
• Use 397 hp at 3,000 rpm only for the reviewed 2016 Silverado HD LML anchor.
• Use 470 hp at 2,800 rpm only for GM's latest reviewed L5P Silverado/Sierra HD test application.
• Keep the current L5P 975-lb-ft torque figure and torque band outside the horsepower field.
• Label the 73-hp and approximately 18.4-percent endpoint differences as arithmetic.
• Do not interpolate missing LB7, LLY, LBZ, LMM, LML or L5P rows.
• Separate factory engine horsepower from measured wheel output.
• Do not infer towing, payload, acceleration, fuel economy, reliability or lifespan.
• Document modifications and calibration separately.
• Use Canadian VIN and recall evidence for Canadian applicability questions.
• Keep inventory, sellers, prices, financing and offers outside the editorial conclusion.

Frequently asked questions

How much horsepower does a 6.6L Duramax have? It depends on the generation and application. The reviewed 2016 Silverado HD LML row is 397 hp at 3,000 rpm; GM's latest L5P Silverado/Sierra HD test row is 470 hp at 2,800 rpm. What changed between those anchors? This review confirms a 73-hp endpoint difference, but it does not attribute the gap to one part or create a complete intervening trend.

Does 470 hp apply to every L5P? No. GM scopes the current result to the latest application it describes and compares it with 2023-and-older L5P engines. Does horsepower establish wheel output, towing, payload, acceleration, reliability, lifespan, Canadian VIN status, candidate condition, inventory, price or offer? No; verify each question with its own evidence.

Editorial conclusion

The defensible answer is a pair of source-scoped anchors, not one universal family number: 397 hp at 3,000 rpm for the reviewed 2016 Silverado HD LML application and 470 hp at 2,800 rpm for GM's latest reviewed L5P Silverado/Sierra HD test application.

Identify the exact engine and truck before choosing a row. Preserve missing generations and applications as unresolved, keep torque and complete-vehicle ratings separate, and treat any current candidate or modification claim as a new evidence problem rather than a brochure lookup.