THE DIESEL DUDES · PUBLISHING DOSSIER · T1a technical

L5P Duramax vs 6.7 Cummins: What Each Manufacturer Actually Publishes

How this article was produced, and how every claim is known to be accurate and sourced. Publishing is paused — this is a dry run.

1. The published article

This is the exact article.html the pipeline produced — the 29-element page that would go live on the blog (brand styling, table of contents, Key Takeaways, the compliance notice before the body, FAQ, numbered Sources, and the machine-readable Article + FAQ data). Publishing is paused; nothing was sent anywhere.

embedded file: article.html · open full page

2. The journey

Every article moves through the same numbered stages. Each move is recorded in the article's manifest.yaml with who did it and when — the article cannot skip a gate.

S1
Research
S2
Calendar
S3
Brief + Outline
S4
Draft + Claims
S5
Editorial QA
S6
Media
S7
Rewrite
S8
Render + Review
MoveByWhen (UTC)What happened
(created) → 01-queuedorchestrator2026-08-13T13:01:08
01-queued → 02-briefingbrief-agent2026-08-14T09:53:263.1-3.11 outline built and frozen
02-briefing → 03-draftingwriter-agent2026-08-14T09:54:15stage S4

source: manifest.yaml

3. QA Gate 1 — the outline, before a word was written

Before drafting, five independent checks run on the outline. This is where the system refuses to build on an unsupported claim. Each check reports how many items it looked at, how many failed, and out of how many (checked / failed / denominator).

Attempt 1 PASS · 2026-08-14T09:53:26
CheckResultchecked/failed/denomDetail
product-accuracyPASS8/0/8tier T1a: 0 product-class claim(s) among 8 declared; 0 SSOT records frozen and available as the enum
legal-frame-and-positionPASS9/0/98 claim frames + 1 triggered elements; ordering sub-check not applicable: outline plans 0 legal and 0 commercial section(s); an ordering constraint ne
citation-entails-claimPASS8/0/88 of 8 planned claims carry an assigned frozen source; 0 declared without one; 0 P3 claim(s) resolved INSUFFICIENT_EVIDENCE and are REMOVED from the o
competitorPASS8/0/88 source/link origins resolved against 10 blocked competitor domains and 4 recognised market-context domains
structure-and-link-livenessPASS6/0/65 sections + 1 required elements; link-liveness sub-check not applicable: the outline plans zero outbound links (predicate `planned_links is non-empty

source: outline-qa-log.json

4. QA Gate 2 — the finished draft

After the full article is written, five verifiers re-check the finished draft — product accuracy, legal accuracy, citations, competitor safety, and structure — and an adjudicator decides the outcome. In this zero-cost run the legal and citation checks are deterministic (exact-match against the sources), which is stronger than an AI opinion and needs no paid API.

VerifierResultchecked/failed/denomDetailFrom
productPASS42/0/42tier T1a: 0 product-class claim(s) of 42 declared; product claims resolve to catalog/ssot/ssot.jsonclaims.json
legalPASS0/0/0deterministic pool match: 0 SUPPORTED, 0 CONTRADICTED, 0 INSUFFICIENT, 0 HUMAN_REVIEWlegal-verify-receipt.json
citationFAIL11/4/117/11 citation gates PASS; 7 sources / 42 placements / reuse 6.0render-receipt.json
competitorPASS8/0/87 cited source host(s) + body scanned against 10 blocked competitor domain(s); hits=nonedraft.md + competitor-allowlist.json
structurePASS25/18/257/25 structural elements present; 2985 rendered words vs 2500 floor (delivery-gate measure; draft body 2670w)render-receipt.json + article.html (delivery-gate word count)
Adjudicator: NEEDS-FIX — citation gate(s) flagged: ['C-1 distinct sources 8-12', 'C-2 placements 18-34', 'C-3 reuse ratio 1.8-3.5', 'C-6 tier mix (>=3 in T1-3, >=2 T4)']
word floor: 2985 rendered words (delivery-gate measure; draft body 2670w) vs 2500 · source: stage-report-S5-S7.json + editorial-review.json

5. Claim-accuracy table — every claim, traced

This is the heart of the proof. Every checkable claim in the article is listed with: the single-source-of-truth it draws on (LEGAL-SSOT = the legal folder, PRODUCT-SSOT = the product catalogue, OEM-TECHNICAL = manufacturer spec sheets), the source's own words (“before”), the verdict, and how it was checked. “After” — our paraphrase — is the claim text itself in column 2.

0
Supported
0
Contradicted
0
Human review
42
Total claims
Claim / verdictOur text (after)ValueSingle source of truthSource's own words (before)Method
c-001
—
GM publishes the L5P rating with its engine speed attached: HORSEPOWER: 470 HP (350 KW) @ 2800 RPM*5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-002
—
The torque figure is given the same way: TORQUE: 975 LB-FT @ 1600 RPM*975 OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-003
—
GM's own summary of the engine puts both together: L5P is even more capable with an impressive power rating of 470 horsepower and 975 lb.-ft. of torque5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-004
—
The asterisk on those numbers is not decoration, and GM says what it covers: *As tested in the Chevrolet Silverado Heavy Duty and GMC Sierra Heavy Duty.—OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-005
—
For the 6.7L Cummins, the current pickup rating on the latest-generation engine page reads: Power 430 hp6.7OEM-TECHNICAL
oem-02
“”Frozen source span (byte-exact)
c-006
—
And the torque figure: Torque 1075 lb-ft1075 lb-ftOEM-TECHNICAL
oem-02
“”Frozen source span (byte-exact)
c-007
—
GM publishes the L5P at Displacement: 6604 CC (403 CI)5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-008
—
Ram's specification sheet for the Cummins at its 2007 introduction gives 408 cubic inches, or about 6,690 cubic centimetres.2007 OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-009
—
GM lists the L5P at Bore x Stroke: 103 x 99 mm5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-010
—
Cummins gives the 6.7L bore as 107 mm, or 4.21 in.6.7OEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-011
—
Its stroke is considerably longer: 124 mm, or 4.88 in.124 OEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-012
—
GM publishes the L5P's peak torque at 1600 rpm and its peak power at 2800 rpm, and it goes one better than a peak by publishing the width of the plateau: ninety percent of pea5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-013
—
GM also states where the engine stops: Maximum Powered Speed: 3450 rpm3450 rpmOEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-014
—
At its introduction, the 6.7L Cummins was published at 350 horsepower at 3,000 rpm and 650 lb.-ft. of torque at 1,500 rpm — replacing the 5.9-litre engine that came before it.6.7OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-015
—
The same specification sheet gives the ceiling: maximum high-idle engine speed of 3,500 rpm.3,500 rpmOEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-016
—
GM publishes a second speed limit specifically for it: Maximum Braking Speed: 4800 rpm4800 rpmOEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-017
—
At the engine's introduction an integrated exhaust brake is available direct from the factory.—OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-018
—
At the engine's introduction an integrated exhaust brake is available direct from the factory.—OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-019
—
The hardware doing it is a variable-geometry unit, described at launch as follows: VGT utilizes 16 fixed vanes and a sliding yoke16 OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-020
—
On the Cummins side the top rating is not available across the range: This high-output (HO) rating is offered exclusively in the RAM 3500 equipped with the Aisin transmission optio3500 OEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-021
—
The volume configuration is rated lower: For RAM 2500 and 3500 (equipped with a 68RFE Auto) HP and torque have increased to 370 hp and 850 lb-ft to provide the right power for thes2500 OEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-022
—
And the top figure itself is a transmission-specific one: The new high-output engine boasts 420 hp and 1,075 lb-ft of best-in-class torque.420 hpOEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-023
—
A fourth rating exists for the commercial chassis: In 2019, RAM 3500, 4500 and 5500 Chassis Cab models equipped with a Aisin automatic, see an increase in horsepower and torque to 2019, OEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-024
—
The transmission is doing more of that work than people expect, and the launch release said as much about the gearbox it arrived with: The new six-speed automatic transmission feat5.16OEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-025
—
The 2019 revision changed the block itself: A Compacted Graphite Iron (CGI) material for the block allows us to increase the cylinder pressure capability while reducing the overall2019 OEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-026
—
The same revision raised injection pressure with a new 29,000 psi (2,000 bar) fuel pump.29,000 psiOEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-027
—
The class threshold it crossed on the way is worth stating: The 2019 model year introduced HD pickup buyers to 1000 lb-ft of torque for the first time ever.2019 OEM-TECHNICAL
oem-06
“”Frozen source span (byte-exact)
c-028
—
GM frames its current engine the same way, against its own earlier versions rather than against a rival, describing the latest L5P as more efficient compared to the previous 2023 a5OEM-TECHNICAL
oem-27
“”Frozen source span (byte-exact)
c-029
—
At launch Ram published life-to-major overhaul intervals of 350,000 miles.350,000 milesOEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)
c-030
—
The 2021 pickup engine is listed at Engine Wet Weight 1,070 lbs2021 OEM-TECHNICAL
oem-01
“”Frozen source span (byte-exact)
c-031
—
The 2019 chassis-cab version of the same engine family is listed at Engine Wet Weight 1026 lb (465 kg)2019 OEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-032
—
The first removed a recurring labour cost: 2019's Hydraulic Lash Adjusters completely eliminate the need to adjust the overhead.2019OEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-033
—
The second stretched a consumable: The interval for the Fleetguard breather filter has been extended to 75,000 miles.75,000 milesOEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-034
—
The fuel side got attention in the same revision, and it is worth knowing what the filtration on the truck actually is: the new fuel-delivery system uses Cummins Filtration “filter—OEM-TECHNICAL
oem-04
“”Frozen source span (byte-exact)
c-035
—
Its published power rating is HORSEPOWER: 305 HP (227 kW) @ 3750 RPM*305 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-036
—
Its torque rating is TORQUE: 495 LB-FT (671 NM) @ 2750 RPM*495 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-037
—
It is also a much smaller engine than the name suggests to anyone thinking in heavy-duty terms: Displacement: 2999 CC (183 CI)2999 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-038
—
Its cylinder dimensions are undersquare in the other direction, at Bore x Stroke: 84 x 90 mm84 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-039
—
And its rev ceiling tells you what it was designed for: Maximum Engine Speed: 5000 rpm5000 rpmOEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-040
—
It is a good engine in that role, and GM's own description of it is worth noting for anyone who assumes the layout follows the badge: The LZ0 Duramax 3.0L turbo diesel is the 0 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-041
—
It also delivers its torque low, with 94 percent of that max twisting force available by 1,500 rpm.94 OEM-TECHNICAL
oem-26
“”Frozen source span (byte-exact)
c-042
—
The same release put a design life on the emissions hardware as well: The emissions system is designed for a government-certified 120,000 miles.120,000 milesOEM-TECHNICAL
oem-23
“”Frozen source span (byte-exact)

source: claims.json (+ legal-verify-receipt.json for legal verdicts)

6. Citation provenance — the paper trail

Each numbered reference in the article traces back to a single source of truth and, from there, to the underlying .gov statute or manufacturer page. A source is linked once and reused by number thereafter (the hyperlink-once rule).

0
Legal claims → legal SSOT
0
Product claims → product SSOT
42
Technical claims → OEM specs
7
Distinct sources (0 legal / 7 OEM/KB / 0 product)
#SSOTTitleTierUsesUnderlying source
[1]OEM-TECHNICALL5p diesel enginetier 210×https://poweredsolutions.gm.com/products/engines/l5p-diesel-engine
[2]OEM-TECHNICAL67l engine ram 2500 and 3500tier 22×https://www.cummins.com/engines/pickup-truck/67l-engine-ram-2500-and-3500
[3]OEM-TECHNICALNewsrelease.dotier 29×https://media.stellantisnorthamerica.com/newsrelease.do?id=6229&mid=97
[4]OEM-TECHNICALCummins 67l turbo diesel 2021tier 26×https://www.cummins.com/engines/products/cummins-67l-turbo-diesel-2021
[5]OEM-TECHNICAL67l cummins turbo diesel 2019 chassis cabtier 27×https://www.cummins.com/engines/products/67l-cummins-turbo-diesel-2019-chassis-cab
[6]OEM-TECHNICALPickup trucktier 21×https://www.cummins.com/engines/pickup-truck
[7]OEM-TECHNICALLz0 enginetier 27×https://poweredsolutions.gm.com/products/engines/lz0-engine

source: render-receipt.json + claims.json by_type

7. Waivers — what the engine refused to invent

When a required part of an article has no source behind it, the engine does not write it anyway. It records a waiver: the element, the reason, and exactly which sources were checked and found empty. This is the rule that would have prevented the August 2026 incident — an element with nothing behind it is declared missing, never filled in with something plausible. Below is every waiver on this article, so the gap is a decision you can see and overrule rather than a silence you would have to notice.

Element 3 — Key Takeaways
no key_takeaways in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:key_takeaways
  • metadata.json:keyTakeaways
Decided 2026-08-14 17:01:10 UTC
Element 4 — Introduction
no intro in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:intro
Decided 2026-08-14 17:01:10 UTC
Element 20 — Vehicle compatibility table
no product on this article carries a fitment record in the catalogue
Sources checked and found empty:
  • catalogue:products[].fitment
Decided 2026-08-14 17:01:10 UTC
Element 22 — Technical specs table
no product on this article carries variant records in the catalogue
Sources checked and found empty:
  • catalogue:products[].variants
Decided 2026-08-14 17:01:10 UTC
Element 8 — Expert quote
no claim on this article is declared as a quote, and the profile's technical-sources pool is a domain allowlist that carries no quotations, so no sourceable attributed quote exists
Sources checked and found empty:
  • <article>/claims.json:type=quote
  • <article>/claims.json:evidence.kind=quote
  • profiles/<client>/technical-sources.json
Decided 2026-08-14 17:01:10 UTC
Element 9 — FAQ
no faq in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:faq
Decided 2026-08-14 17:01:10 UTC
Element 16 — The Bottom Line
no bottom_line in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:bottom_line
Decided 2026-08-14 17:01:10 UTC
Element 24 — Internal link block
no internal_links in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:internal_links
Decided 2026-08-14 17:01:10 UTC
Element 27 — Named author bio
no author_bio/credentials in metadata (author in 2b)
Sources checked and found empty:
  • metadata.json:author_bio
Decided 2026-08-14 17:01:10 UTC
Element 25 — Entity optimization block
no product entities on this article and no publisher entity to state
Sources checked and found empty:
  • profile.yaml:display_name
  • <article>/callouts.json
Decided 2026-08-14 17:01:10 UTC
Element 26 — Owned data / customer review block
no product referenced by this article carries a review aggregate in the catalogue, and there is no review text anywhere in it to quote
Sources checked and found empty:
  • catalogue:products[].reviews.rating
  • catalogue:products[].reviews.count
Decided 2026-08-14 17:01:10 UTC
Element 11 — AI block (GEO)
the block reuses the Key Takeaways bullets and none exist (element 3 is itself waived)
Sources checked and found empty:
  • metadata.json:key_takeaways
  • metadata.json:keyTakeaways
Decided 2026-08-14 17:01:10 UTC
Element 23 — Visible update date
metadata carries no date_modified, and the render clock is not a substitute for a recorded edit date
Sources checked and found empty:
  • metadata.json:date_modified
Decided 2026-08-14 17:01:10 UTC

source: waivers.json · cross-checked against the 13 waiver notice(s) rendered in article.html

8. Copied word-for-word? — overlap with the sources

A paraphrase is supposed to differ from its source — that is what makes it a paraphrase. This measures the opposite: how many of the article's words sit inside a run of 8 or more consecutive words that also appears, exactly, in one of the frozen sources the writer worked from. This is shown, never enforced. It cannot block or delay anything, because some overlap is correct — legal wording has to track the statute closely, and a model name or a torque figure has one right spelling. Read a high number as "go and look", not as "this is wrong".

13.1%
Words inside a copied run
335
of 2562 words
29
Longest copied run (words)
41
Sources compared
Low — consistent with paraphrasing, not copying · flag threshold 30%, a review trigger rather than a limit — nothing is blocked at any value

The longest copied runs, and where each came from:

LengthSourceThe copied words
29woem-01“For RAM 2500 and 3500 (equipped with a 68RFE Auto) HP and torque have increased to 370 hp and 850 lb-ft to provide the right power for these models.”
28woem-04“In 2019, RAM 3500, 4500 and 5500 Chassis Cab models equipped with a Aisin automatic, see an increase in horsepower and torque to 360 hp and 800 lb-ft.”
27woem-04“A Compacted Graphite Iron (CGI) material for the block allows us to increase the cylinder pressure capability while reducing the overall weight by almost 60 lbs. The”
21woem-23“Testing has shown more than a 30 percent improvement in retarding torque at 2000 rpm compared with traditional brake exhaust methods.”
18woem-27“L5P is even more capable with an impressive power rating of 470 horsepower and 975 lb.-ft. of torque”
18woem-01“This high-output (HO) rating is offered exclusively in the RAM 3500 equipped with the Aisin transmission option. The”
18woem-06“The 2019 model year introduced HD pickup buyers to 1000 lb-ft of torque for the first time ever.”
17woem-26“The LZ0 Duramax 3.0L turbo diesel is the second generation inline-six turbo diesel over the previous LM2.”
14woem-23“The new six-speed automatic transmission features a best-in-class gear ratio spread of 5.16:1”
14woem-04“2019's Hydraulic Lash Adjusters completely eliminate the need to adjust the overhead. The”
13woem-27“As tested in the Chevrolet Silverado Heavy Duty and GMC Sierra Heavy Duty.”
13woem-27“ninety percent of peak torque is produced from 1,550 rpm thru 2,850 rpm”

Words copied, by source:

SourceWords
oem-0498 words
oem-2377 words
oem-0160 words
oem-2753 words
oem-2629 words
oem-0618 words
Read the draft with the copied runs highlighted — every highlight names the source it matches
# L5P Duramax vs 6.7 Cummins: What Each Manufacturer Actually Publishes Most versions of this comparison quote numbers with no source and no configuration attached. This one uses what GM and Cummins publish on their own engine pages, plus Ram's specification sheet for the Cummins, and it names the engine speed each figure arrives at — because on a towing engine, where the peak lands matters as much as how big it is. Two things to keep in mind while reading. Each figure belongs to a specific truck and drivetrain rather than to an engine in the abstract, and figures from different model years describe genuinely different hardware. Both cautions get their own section below, because most disagreements about these two engines come down to one of them. ## The headline numbers, from each manufacturer GM publishes the L5P rating with its engine speed attached: HORSEPOWER: 470 HP (350 KW) @ 2800 RPM* The torque figure is given the same way: TORQUE: 975 LB-FT @ 1600 RPM* GM's own summary of the engine puts both together: L5P is even more capable with an impressive power rating of 470 horsepower and 975 lb.-ft. of torqueoem-27 The asterisk on those numbers is not decoration, and GM says what it covers: *As tested in the Chevrolet Silverado Heavy Duty and GMC Sierra Heavy Duty.oem-27 For the 6.7L Cummins, the current pickup rating on the latest-generation engine page reads: Power 430 hp And the torque figure: Torque 1075 lb-ft So on published peaks the Duramax leads on horsepower and the Cummins leads on torque, by roughly forty of one and a hundred of the other. That is the whole of what the headline numbers tell you, and it is less than most comparisons claim from them. ## Displacement: the badges are almost a coincidence The two engines are sold as a 6.6 and a 6.7, which sounds like a tenth of a litre between them. In cubic inches the gap is smaller still. GM publishes the L5P at Displacement: 6604 CC (403 CI) Ram's specification sheet for the Cummins at its 2007 introduction gives 408 cubic inches, or about 6,690 cubic centimetres. Five cubic inches apart, out of more than four hundred. Whatever separates these engines in behaviour, capacity is not it — which is worth establishing early, because "bigger engine" is the most common shorthand in this argument and it is not describing anything real. ## Bore and stroke: where the two designs actually part company The dimensions of each cylinder tell you far more than the total, and both manufacturers publish them. GM lists the L5P at Bore x Stroke: 103 x 99 mm Cummins gives the 6.7L bore as 107 mm, or 4.21 in. Its stroke is considerably longer: 124 mm, or 4.88 in. Read those together and the two design philosophies separate cleanly. The Duramax's bore and stroke are within four millimetres of each other — near enough to square, a shape that breathes well and is happy turning higher revolutions. The Cummins' stroke is seventeen millimetres longer than its bore, which is what engineers call undersquare: the piston travels further per rotation and pushes on the crankshaft through a longer lever. That single difference predicts most of what the headline numbers then show. A longer stroke produces more turning effort at low engine speed and runs out of breath earlier; a squarer engine gives up some of that low-speed effort and keeps making power further up the range. The Duramax leading on horsepower while the Cummins leads on torque is not a coincidence of tuning. It is the geometry showing through in the results. ## Where the peaks arrive, and how wide the plateau is The more useful comparison is engine speed, because it describes how each engine behaves with a trailer behind it. GM publishes the L5P's peak torque at 1600 rpm and its peak power at 2800 rpm, and it goes one better than a peak by publishing the width of the plateau: ninety percent of peak torque is produced from 1,550 rpm thru 2,850 rpmoem-27 That is the most practically useful number on either page. A 1,300 rpm window holding ninety percent of maximum torque means the engine does not need to be kept in a narrow band to pull well, which is what actually matters when a grade steepens and the transmission has to choose a gear. GM also states where the engine stops: Maximum Powered Speed: 3450 rpm The Cummins figures that carry an engine speed come from Ram's launch specification for the engine, which describes the hardware as introduced rather than a truck on a lot today. At its introduction, the 6.7L Cummins was published at 350 horsepower at 3,000 rpm and 650 lb.-ft. of torque at 1,500 rpmoem-23 — replacing the 5.9-litre engine that came before it. The same specification sheet gives the ceiling: maximum high-idle engine speed of 3,500 rpm. Those two ceilings — 3450 rpm and 3,500 rpm — are close enough to be treated as the same answer, and it is a low one by petrol standards. Both of these engines do their work in the bottom half of a rev range that a gasoline V8 would consider idle-adjacent. The torque peaks, at 1600 rpm for the current L5P and 1,500 rpm for the Cummins as introduced, are close as well. Where the engines genuinely differ is above those peaks, and that is the part of the curve that gets you up a long grade without a downshift. ## Engine braking is half of towing, and both pages address it Going down a grade with weight behind you asks a different question of an engine, and each manufacturer answers it in its own vocabulary. GM publishes a second speed limit specifically for it: Maximum Braking Speed: 4800 rpm That figure is far above the powered limit because the engine is being driven by the truck rather than driving it, and it tells you how much retarding work the hardware is rated to absorb. Ram approached the same job through the turbocharger. At the engine's introduction an integrated exhaust brake is available direct from the factory.oem-23 The launch release published what that bought: Testing has shown more than a 30 percent improvement in retarding torque at 2000 rpm compared with traditional brake exhaust methods.oem-23 The hardware doing it is a variable-geometry unit, described at launch as follows: VGT utilizes 16 fixed vanes and a sliding yokeoem-23 Two different framings of one problem. GM gives a rated speed for the braking mode; Ram gives an improvement in braking torque at a working engine speed. Neither converts into the other, which is why descent performance is worth a test drive on a real hill rather than a spec-sheet argument. ## Each figure is a configuration rating, not an engine's identity This is the caution that sinks most online comparisons, and it applies to both brands. On the Cummins side the top rating is not available across the range: This high-output (HO) rating is offered exclusively in the RAM 3500 equipped with the Aisin transmission option. Theoem-01 volume configuration is rated lower: For RAM 2500 and 3500 (equipped with a 68RFE Auto) HP and torque have increased to 370 hp and 850 lb-ft to provide the right power for these models.oem-01 And the top figure itself is a transmission-specific one: The new high-output engine boasts 420 hp and 1,075 lb-ft of best-in-class torque.oem-01 A fourth rating exists for the commercial chassis: In 2019, RAM 3500, 4500 and 5500 Chassis Cab models equipped with a Aisin automatic, see an increase in horsepower and torque to 360 hp and 800 lb-ft.oem-04 Four published numbers, one engine. The transmission is doing more of that work than people expect, and the launch release said as much about the gearbox it arrived with: The new six-speed automatic transmission features a best-in-class gear ratio spread of 5.16:1oem-23 A wide ratio spread lets the same engine cover a launch gear and a highway overdrive, which is why a rating quoted without its transmission is only part of a description. GM's asterisk does the same job from the other direction, naming the two trucks the rating was measured in. So a Ram 2500 owner and a Ram 3500 owner can quote different Cummins figures and both be reading the manufacturer correctly, while a Silverado owner is quoting a number measured in their own truck. Comparing one manufacturer's best rating against another's mid rating produces whatever answer you started with. Any comparison worth anything states the chassis and the transmission alongside the number. **A peak is also not a duty cycle.** Sustained load is a different question from a maximum, and it is the one a work truck asks every day. Both of these engines are rated for heavy-duty pickups, and a peak figure does not settle which of them suits a given job. ## Both engines have moved, and not from the same starting point The Cummins' current rating is a long way from where the 6.7 began, and that context is easy to lose in a comparison of today's numbers. Set the launch figures against the current rating and the scale of the change is clear: torque is up by more than four hundred lb-ft over the engine's life, on the same displacement and the same bore and stroke. Some of how it got there is published. The 2019 revision changed the block itself: A Compacted Graphite Iron (CGI) material for the block allows us to increase the cylinder pressure capability while reducing the overall weight by almost 60 lbs. Theoem-04 same revision raised injection pressure with a new 29,000 psi (2,000 bar) fuel pump.oem-04 Higher cylinder pressure capability and higher injection pressure are how a diesel gains output without gaining size, and both of those changes are the reason a modern 6.7 is a different engine from a 2007 one wearing the same badge. The class threshold it crossed on the way is worth stating: The 2019 model year introduced HD pickup buyers to 1000 lb-ft of torque for the first time ever.oem-06 GM frames its current engine the same way, against its own earlier versions rather than against a rival, describing the latest L5P as more efficient compared to the previous 2023 and older L5P engines.oem-27 Which means a comparison assembled from figures found in two different years is not comparing these two engines. It is comparing one manufacturer's current hardware against another manufacturer's history. ## A published rebuild interval One durability figure is stated outright rather than implied. At launch Ram published life-to-major overhaul intervals of 350,000 miles. That is a design interval to a major overhaul rather than a coverage promise, and it is not a prediction about any particular truck. It is still a useful number, because it tells you what the engine's bottom end was engineered around, and it is the sort of figure that separates a heavy-duty diesel from a light-duty one. The same release put a design life on the emissions hardware as well: The emissions system is designed for a government-certified 120,000 miles.oem-23 Those two numbers describe different things and are worth keeping apart. One is when the engine is expected to want a rebuild; the other is what the aftertreatment was certified to survive, and it is the shorter of the two by a wide margin. ## What the pages say about living with each engine Peak figures are the least of what an engine costs its owner, and the Cummins pages carry a few numbers that bear on that directly. Weight is one of them, and it is published per application rather than once. The 2021 pickup engine is listed at Engine Wet Weight 1,070 lbs The 2019 chassis-cab version of the same engine family is listed at Engine Wet Weight 1026 lb (465 kg) Several hundred pounds of engine over the front axle is a real part of how a heavy-duty truck rides and steers unloaded, and it is one of the trade-offs that comes with a long inline-six rather than a shorter engine. Two service items changed in a way owners notice. The first removed a recurring labour cost: 2019's Hydraulic Lash Adjusters completely eliminate the need to adjust the overhead. Theoem-04 second stretched a consumable: The interval for the Fleetguard breather filter has been extended to 75,000 miles.oem-04 Both of those are the kind of change that never appears in a horsepower comparison and shows up on every invoice. Valve-lash adjustment on a six-cylinder diesel is a labour job, and removing it from the schedule removes a whole shop visit from the ownership arc. The fuel side got attention in the same revision, and it is worth knowing what the filtration on the truck actually is: the new fuel-delivery system uses Cummins Filtration “filter in filter” NanoNet® technology. ## The 3.0 Duramax is a different class One clarification, because the name causes genuine confusion: GM's 3.0-litre Duramax is not part of this comparison. Its published power rating is HORSEPOWER: 305 HP (227 kW) @ 3750 RPM* Its torque rating is TORQUE: 495 LB-FT (671 NM) @ 2750 RPM* It is also a much smaller engine than the name suggests to anyone thinking in heavy-duty terms: Displacement: 2999 CC (183 CI) Its cylinder dimensions are undersquare in the other direction, at Bore x Stroke: 84 x 90 mm And its rev ceiling tells you what it was designed for: Maximum Engine Speed: 5000 rpm That ceiling is around fifteen hundred rpm above either heavy-duty engine's limit, which is the clearest single indication that this is a light-duty half-ton engine rather than a small heavy-duty one. It is a good engine in that role, and GM's own description of it is worth noting for anyone who assumes the layout follows the badge: The LZ0 Duramax 3.0L turbo diesel is the second generation inline-six turbo diesel over the previous LM2.oem-26 It also delivers its torque low, with 94 percent of that max twisting force available by 1,500 rpm.oem-26 So a "Duramax vs Cummins" argument that quotes 3.0-litre figures against heavy-duty Cummins figures is comparing different classes of truck, and usually different jobs entirely. ## The honest summary On the numbers each manufacturer publishes: the L5P leads horsepower, the 6.7L leads torque, the L5P's torque peak is published at a low 1600 rpm with ninety percent of it held across a 1,300 rpm plateau, and the two engines are within five cubic inches of the same size. The Cummins is the longer-stroke design and its figures follow the chassis and transmission it is fitted to; the Duramax is the squarer one and its figures were measured in the Silverado and Sierra Heavy Duty. Anything more specific than that — sustained-load behaviour, real-world towing, longevity on a particular truck — is decided by how the truck is used rather than by either engine page. --- ## Sources 1. GM Powered Solutions, L5P 6.6L Duramax diesel engine — https://poweredsolutions.gm.com/products/engines/l5p-diesel-engine 2. GM Powered Solutions, LZ0 3.0L Duramax engine — https://poweredsolutions.gm.com/products/engines/lz0-engine 3. Cummins Inc., 6.7L Turbo Diesel for Ram 2500 and 3500 — latest generation engine page 4. Cummins Inc., Cummins 6.7L Turbo Diesel (2021) and 6.7L Cummins Turbo Diesel (2019) for Chassis Caboem-04 engine pages 5. Cummins Inc., pickup truck engine overview 6. Stellantis North America media newsroom, 6.7L Cummins introduction release and specifications — https://media.stellantisnorthamerica.com/newsrelease.do?id=6229&mid=97

source: lib/overlap over draft.md vs frozen/sources/*.txt · measured on the authored draft, not the rendered page, so the legal notices and catalogue product text — which are verbatim by design — cannot inflate it

9. Verdict