Knife Steel Comparison Tool: Compare Any 2-4 Steels
Compare two to four knife steels side by side. At equal hardness CPM 3V is more than nine times tougher than M390, and every number is sourced.
Data updated August 2026 · 25 steels · every number traced to a published test
At the same hardness, CPM 3V is more than nine times tougher than M390 — 41.5 ft-lb at 61.5 HRC against 4.5 ft-lb at 61.9 HRC. Almost every knife steel chart on the internet compresses that into “3V: 5 dots, M390: 2 dots.” This tool does not. It shows you the actual published test numbers, the hardness each one was measured at, and it tells you plainly when a number has never been published rather than inventing a dot for it.
Pick two to four steels and you get their measured impact toughness, their measured CATRA edge retention, their corrosion rating, their full published chemistry and the knives they actually ship in. Then pick how you use a knife and the tool ranks every steel that has complete data, using arithmetic you can check, printed on the page.
- Toughness: how hard it is to chip or snap
- Stays sharp: how long the edge lasts
- Rust resistance: 0 to 10
- No test result published
Pick steels above to compare them.
That line gives us credit and helps people find the full tool.
Sign up for the free Popular EDC email. We will tell you when this tool gets new steels or new test results.
Where these numbers come from
Three of the four things people argue about — toughness, edge retention, corrosion resistance — have actually been measured, by one person, on comparable samples. Larrin Thomas, a metallurgist who wrote the book on knife steel, runs unnotched Charpy impact tests and CATRA edge-retention tests and publishes the results. Almost nobody quotes his numbers. They quote his conclusions, convert them into five-dot ratings, and lose the part that matters.
Toughness is subsize unnotched Charpy impact energy in foot-pounds, three or more specimens averaged. Edge retention is CATRA total cards cut, reported in millimeters of card stock, with the blade sharpened to 15 degrees per side on a 400 grit CBN wheel. Corrosion resistance is a 0 to 10 rating from salt-spray panels, where 10 is a steel like Vanax or LC200N that does not rust in 3.5 percent saltwater and 0 is 1095. Chemistry comes from the maker’s own datasheet — Crucible and Niagara, Böhler-Uddeholm, Carpenter, Alleima, and for two budget steels a compiled nominal composition because no maker publishes one.
The number beside every toughness value is the hardness it was measured at, and you should not ignore it. Toughness collapses as hardness climbs; the same steel run two points harder can lose a third of its impact energy. Charts that give you one dot per steel are hiding a variable that changes the answer. That is why this tool prints the hardness next to the number every single time, and why the 420HC result below is not a typo.
How the ranking arithmetic works
Each of the three measured axes is normalized to a 0 to 100 scale across the sixteen steels that have all three numbers published: score = 100 × (value − lowest) ÷ (highest − lowest). For “easy to sharpen” the edge-retention axis is inverted — 100 − normalized edge retention — because wear resistance is precisely the property that makes a steel slow on a stone. Those three normalized numbers are then combined with the weights printed above the results, and the total is divided by 100 so the output reads 0 to 100. There is no fudge factor and no editorial thumb on the scale. Change the use case and you change only the weights.
A worked example
Take S30V against S35VN, the comparison people make more than any other. S30V measures 5.5 ft-lb at 60.0 HRC and 596 mm CATRA at 61.9 HRC. S35VN measures 9.5 ft-lb at 60.9 HRC and 520 mm CATRA at 61.2 HRC. So S35VN is not “a bit tougher”: it is 73 percent tougher at essentially the same hardness, and it pays for that with 13 percent less edge retention. Both carry the same 7.5 out of 10 corrosion rating. If you cut cardboard all day, the 13 percent matters more; if you have ever chipped a S30V blade, the 73 percent matters more. That is the whole trade, in two real numbers, and no five-dot chart can show it to you.
Four things the measured data says that the dot charts do not.
- The premium stainless steels are, as a class, the least tough steels here. M390 (4.5 ft-lb), CPM S110V (4.0), D2 (5.1), 154CM (5.4), S30V (5.5), VG-10 (5.8) all sit in a tight band near the bottom. CPM 3V sits at 41.5. That is not a small edge in one direction; it is roughly an order of magnitude.
- The two toughest stainless steels in this set are two of the cheapest. Buck’s 420HC measures 40.8 ft-lb at 57.5 HRC — effectively level with CPM 3V — and 14C28N, the steel that shows up on sub-$60 folders, measures 30.3 ft-lb at 60.7 HRC, which beats every powder stainless on this list including CPM MagnaCut at 17.6. If your knife’s job is to survive being used badly, the budget steel is the better engineering answer, and it has been all along.
- Toughness and edge retention really do trade against each other, and we can now put a number on it. Across the sixteen steels with both measurements published, the Spearman rank correlation between impact toughness and CATRA edge retention is −0.70. Strong, negative, and not absolute — which is exactly why the exceptions are interesting.
- CPM CruWear and CPM MagnaCut are the real exceptions. CruWear posts 528 mm of CATRA and 20.1 ft-lb of toughness, beating S35VN on both axes at once. MagnaCut is the only steel here that pairs a corrosion rating above 9 with toughness in double figures; every other 9-and-above steel (M390 at 4.5, S110V at 4.0) is in the bottom four for toughness. Those two are not marketing. They are visible in the data.
The measured data, all 25 steels
Sorted by measured toughness, highest first. Blank cells are honest: they mean the test has never been published for that steel, not that it scored badly. Six steels have no published Charpy number and seven have no published CATRA number, and the tool refuses to rank them rather than guess.
| Steel | Family | Toughness | Edge retention (CATRA) | Corrosion | Typical production HRC | Introduced | Example production knives |
|---|---|---|---|---|---|---|---|
| CPM 3V | powder tool | 41.5 ft-lb | |||||
Published chemistry, all 25 steels
Weight percent, from the maker’s datasheet where one exists. This is the part that explains the performance: vanadium and niobium carbides drive edge retention, chromium in solution drives corrosion resistance, and both of them, in excess, are what cost you toughness.
| Steel | C | Cr | V | Mo | W | Co | N | Mn | Si | Nb | Ni | Cu | S | P |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1095 | 0.95 | — | — | — | — | — | — | 0.4 | — | — | — | — | — | — |
| 14C28N | 0.62 | 14 | — | — | — | — | 0.11 | 0.6 | 0.2 | — | — | — | — | — |
| 154CM | 1.05 | 14 | — | 4 | — | — | — | — | — | — | — | — | — | — |
| 420HC | 0.45 | 13 | — | — | — | — | — | — | — | — | — | — | — | — |
| 8Cr13MoV | 0.8 | 13 | 0.1 | 0.15 | — | — | — | — | — | — | — | — | — | — |
| AEB-L | 0.68 | 12.9 | — | — | — | — | — | 0.6 | 0.4 | — | — | — | — | — |
| AUS-8 | 0.72 | 14 | 0.2 | 0.2 | — | — | — | — | — | — | — | — | — | — |
| CPM 3V | 0.8 | 7.5 | 2.75 | 1.3 | — | — | — | 0.3 | 1 | — | — | — | — | — |
| CPM 4V | 1.35 | 5 | 3.85 | 2.95 | — | — | — | 0.4 | 0.8 | — | — | — | — | — |
| CPM CruWear | 1.15 | 7.5 | 2.4 | 1.6 | 1 | — | — | — | — | — | — | — | — | — |
| CPM MagnaCut | 1.15 | 10.7 | 4 | 2 | — | — | 0.2 | — | — | 2 | — | — | — | — |
| CPM Rex 45 | 1.3 | 4.05 | 3.05 | 5 | 6.25 | 8 | — | — | — | — | — | — | 0.06 | — |
| CPM S110V | 2.8 | 15.25 | 9 | 2.25 | — | 2.5 | — | — | — | 3 | — | — | — | — |
| CPM S90V | 2.3 | 14 | 9 | 1 | — | — | — | — | — | — | — | — | — | — |
| CPM-154 | 1.05 | 14 | — | 4 | — | — | — | — | — | — | — | — | — | — |
| D2 | 1.42 | 11.5 | 0.8 | 0.9 | — | — | — | — | — | — | — | — | — | — |
| Elmax | 1.7 | 18 | 3 | 1 | — | — | — | 0.3 | 0.8 | — | — | — | — | — |
| K390 | 2.47 | 4.2 | 9 | 3.8 | 1 | 2 | — | 0.4 | 0.55 | — | — | — | — | — |
| M390 / 20CV / 204P | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
| Maxamet | 2.15 | 4.75 | 6 | — | 13 | 10 | — | 0.3 | 0.25 | — | — | — | 0.07 | — |
| Nitro-V | 0.68 | 13 | 0.08 | — | — | — | 0.11 | 0.65 | 0.4 | — | — | — | — | — |
| S30V | 1.45 | 14 | 4 | 2 | — | — | — | — | — | — | — | — | — | — |
| S35VN | 1.4 | 14 | 3 | 2 | — | — | — | — | — | 0.5 | — | — | — | — |
| S45VN | 1.48 | 16 | 3 | 2 | — | — | 0.15 | — | — | 0.5 | — | — | — | — |
| VG-10 | 1 | 15 | 0.25 | 1 | — | 1.55 | — | — | — | — | — | — | — | — |
What to know about each steel
| Steel | Worth knowing | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 1095 | MATERIAL CAVEAT: KSN’s CATRA specimen was heat treated to 61.8 HRC, whereas most production 1095 knives (e.g. ESEE) run 55-58 HRC. | ||||||||
| The published 316 mm figure therefore overstates what a typical 1095 production knife will do. 1095 is a standard AISI grade with no single introduction year and no proprietary datasheet. | |||||||||
| 14C28N | Alleima’s datasheet gives only annealed/cold-rolled hardness in HV/HRB, not a hardened HRC range — hence hrc_typical is [VERIFY] (KSN’s own test specimen was 62 HRC; 58-62 is the common production window but not maker-published). Developed by Sandvik for Kershaw as a corrosion-improved 13C26/AEB-L; Kershaw knives in it were being sold by 2009 (KSN ‘All About AEB-L’). Sandvik’s strip business became Alleima in 2022. | ||||||||
| 154CM | CATRA is [VERIFY]: Larrin’s CATRA program tested CPM-154 (the powder version, 492 mm at 61 HRC) but NOT conventional ingot 154CM. I have deliberately not carried the CPM-154 number over. Developed by Crucible in 1959 with the Wright Air Development Center for jet engine applications; Bob Loveless began using it in knives in 1972. ATS-34 (Hitachi) is the near-identical Japanese equivalent. | ||||||||
| That 420 point appears only in the CATRA scatter chart (approximately 319 mm at 55 HRC by chart reading) and is a different, lower-carbon steel — so catra_tcc for 420HC is [VERIFY] rather than borrowing the 420 number. AISI 420 was first announced in 1915 and was developed for knives (per KSN), but 420HC specifically has no published introduction year. Buck’s public heat-treating page does not state an HRC value for 420HC. | |||||||||
| Its 0-10 ratings are Larrin’s judgment, not measurements, and are shared with AUS-8 (toughness 6, edge retention 3, corrosion 7). No manufacturer datasheet is publicly published; composition above is a compiled nominal. | |||||||||
| AEB-L, 13C26 and (nearly) Nitro-V/14C28N are the same base steel. Toughness is unusually sensitive to austenitizing temperature: 1925F gives ~37.6 ft-lbs, 2025F only ~19.5 ft-lbs at similar hardness. No maker-published hardened HRC range found. | |||||||||
| Larrin explicitly did NOT test AUS-8 for toughness or CATRA — no published numbers exist, so both fields are [VERIFY]. KSN 0-10 ratings (toughness 6, edge retention 3, corrosion 7) are judgment calls shared with 8Cr13MoV. | |||||||||
| CPM 3V | Larrin: patented in 1997, ‘CPM-3V became Crucible’s go-to steel for applications requiring high toughness; they released it around 1999.’ Its ancestor Vasco Die was patented by Harry Johnstin of VASCO in 1964. | Note the CATRA specimen was 60.1 HRC while the 2025 toughness specimen was 61.5 HRC — slightly different hardnesses. Larrin’s 2025 retest found newer 3V material is notably tougher than his older material at the same hardness, so older KSN charts understate it. | |||||||
| Crucible’s own MagnaCut datasheet does publish 4V at 36 ft-lbs @ 62 HRC, but by quarter-size unnotched converted to Charpy C-notch — a different test that is not comparable to Larrin’s subsize unnotched scale, so it cannot substitute. Larrin rates 4V and Uddeholm Vanadis 4 Extra as effectively the same steel and rates them jointly (toughness 7, edge retention 5, corrosion 4). Crucible positions 4V as a higher-wear upgrade to 3V. | |||||||||
| Larrin’s toughness testing was done on Z-Wear and he treats them as the same steel. Conventional ingot Cru-Wear is a materially different product (8.7 vs ~20 ft-lbs) — do not conflate. Larrin’s testing also shows the common 1000F ‘upper temper’ costs roughly 2-5 ft-lbs of toughness versus a 400F temper. Crucible’s datasheet notes CPM CruWear normally tempers at 900-1050F. No published introduction year found for the CPM version; the conventional Cru-Wear descends from Vasco Wear/Vasco Die. | |||||||||
| CPM MagnaCut | CATRA TCC is [VERIFY]: Larrin’s numeric TCC table pre-dates MagnaCut, and MagnaCut appears only as a plotted point on the CATRA scatter chart (its two-point green line reads roughly 505 mm and 568 mm, but I will not report a chart-eyeballed figure as data). Larrin states MagnaCut ‘matched the approximate edge retention of S35VN, CPM-4V, and CPM-CruWear’ (520, 538, 528 mm respectively). Crucible’s datasheet gives CATRA as 135% of 440C at 62.5 HRC (vs S35VN 130% at 61 HRC). Designed by Larrin Thomas himself and announced 25 March 2021 — he is therefore not a neutral party on this steel. Its distinguishing feature is zero chromium carbide, giving stainless corrosion resistance with non-stainless-tool-steel toughness. | ||||||||
| CPM Rex 45 | Both toughness and CATRA are [VERIFY]. Rex 45 was added in Larrin’s May 2021 ‘seven more steels’ article; the published table gives heat treatment and hardness only (2100F / 10 min / plate quench / 12 hr cryo / 1000F temper -> 66.5 HRC), with the TCC value appearing solely as a plotted point. MATERIAL CAVEAT: Larrin deliberately heat treated Rex 45 to the high end (66.5 HRC) because ‘the Spyderco knives produced in the steel are in the upper end of hardness’ — so its chart position is not hardness-matched to the ~61 HRC used for most other steels, and he notes that when compensated for hardness CPM M4 actually has better wear resistance. Sold under multiple names: CPM Rex 45 (Crucible), Erasteel ASP 2030, Uddeholm Vanadis 30, Bohler S590, and Hitachi HAP40 in Japan — Larrin rates Rex 45 and HAP40 jointly. It is a PM version of M3:2 high speed steel plus 8% cobalt. The Crucible datasheet’s hardness tables are aimed at tooling (standard recommendation to reach HRC 66-68), not knives. | ||||||||
| Two distinct published chemistries exist — the one above is the current/revised version. Larrin describes S110V as trading some of S90V’s toughness for better corrosion resistance. Crucible’s datasheet notes it is available only on special request as as-HIP, clad, or hot-rolled sheet. | |||||||||
| CPM S90V | Larrin: ‘CPM-S90V was originally named CPM-420V released around 1996.’ Sold under both names historically. | Note the maker’s aim hardness (56-59) is well below where knife makers actually run it and below the 61.6 HRC at which the CATRA number was measured. Larrin names S90V his favorite of the high-edge-retention stainless steels because of its comparatively decent toughness. | |||||||
| WARNING about older KSN charts: an early CPM-154 toughness measurement of ~18.4 ft-lbs at ~60 HRC circulated on Larrin’s 2020 charts and was later retested and retracted — he found every comparable-carbide stainless maxes out near 10 ft-lbs. Use ~8.5 ft-lbs, not 18. Crucible’s own MagnaCut datasheet separately lists CPM 154 at 26 ft-lbs at 60 HRC using quarter-size unnotched specimens converted to C-notch — a different test, not comparable. | |||||||||
| D2 has ~12% Cr but is NOT stainless because most of the chromium is tied up in chromium carbide. A powder-metallurgy version, CPM-D2, was released around 2007 and has notably better toughness than ingot D2 — the numbers above are for conventional ingot D2. | |||||||||
| The datasheet quotes physical properties ‘hardened and tempered to 58 HRC’ and a delivery condition of ~280 HB soft annealed, but does not publish a hardened working HRC range, so hrc_typical is [VERIFY]. | |||||||||
| The Bohler product page publishes chemistry but no attainable HRC, hence hrc_typical is [VERIFY]. Bohler Microclean powder metallurgy steel; used by Spyderco. No published introduction year located. | |||||||||
| M390 / 20CV / 204P | IMPORTANT toughness correction: earlier KSN charts showed M390 at 8-9 ft-lbs (8.2 and 8.8 @ ~61.8 HRC, 2140F/365F). Larrin retested and concluded those specimens carried excess retained austenite from too-high an austenitizing temperature; he does NOT recommend that heat treat and the corrected value is ~4.5 ft-lbs. Use 4.5, not 8-9. Patent date is stated two ways by Larrin: the M390 history article says the patent application was submitted in 1989, the S90V article says Bohler patented M390 in 1988 — both are cited here rather than picking one. Commercial knife timeline: Latrobe Duratech 20CV appeared on Timken’s site in 2005; Carpenter introduced ‘Micromelt 20-4’ in 2009, renamed CTS-204P by early 2010; Crucible began selling CPM-20CV after Carpenter bought Latrobe in 2011. The three brands are treated as one steel by Larrin’s ratings. | ||||||||
| Maxamet | Both toughness and CATRA are [VERIFY]. Maxamet was added in Larrin’s May 2021 ‘seven more steels’ article, whose published table gives only heat treatment and hardness (2100F / 10 min / plate quench / 12 hr cryo / 1000F temper -> 68.6 HRC) — no TCC column. Its TCC exists only as a point on the CATRA scatter chart, and its toughness only as a point in the densest region of the high-alloy toughness chart. The one numeric statement Larrin makes is qualitative-comparative: ‘At 66.7 Rc it had toughness equal to 15V at 65.3 Rc. At 68+ Rc the toughness becomes somewhat less impressive.’ KSN rates its edge retention 11 on a nominally 1-10 scale (he deliberately exceeds 10 for Maxamet and Rex 121 to avoid compressing everything else) and toughness 2. CAUTION for a comparator: KSN’s specimens were 66.7-68.6 HRC, far above most steels’ test hardness, so cross-steel comparison at equal hardness is not valid. US Patent 6,482,354. | ||||||||
| NJSB publishes a heat-treat PDF but no headline HRC range on the product page, so hrc_typical is [VERIFY]. Larrin notes the 0.08% V is too little to form vanadium carbide and contributes essentially nothing to wear resistance. | |||||||||
| Larrin separately found CPM SPY27 behaves almost identically to S35VN in both toughness and edge retention. | |||||||||
| S45VN | Crucible’s S45VN datasheet is headed ‘New for fall 2019’; Larrin’s launch article is dated 1 November 2019. | Designed as a corrosion- and wear-improved S35VN using more chromium plus nitrogen and niobium. Crucible’s MagnaCut datasheet lists S45VN at 19 ft-lbs @ 61.5 HRC and 140% of 440C for CATRA — different test methods from the KSN numbers above. | |||||||
| Takefu notes its published chemistry values are ‘representative data for the standard values of each steel grade’. The cobalt addition is unusual for a stainless and was likely for tempering resistance rather than hardness. |
The thing this tool cannot tell you
Heat treatment matters more than the steel on the label, and there is no public dataset for it. Every number above comes from a specimen heat treated by a metallurgist to get the best out of that alloy. The folder in your pocket was heat treated by a factory to hit a price and a schedule. A well-run 14C28N blade will beat a badly-run MagnaCut blade, and you cannot tell which you have from the steel stamp. When two knives use the same steel and one costs three times as much, heat treatment and geometry are usually where the money went.
Geometry beats steel for how well a knife cuts. A thin, well-ground blade in a mediocre steel will out-slice a thick one in a super steel, and CATRA testing deliberately standardises geometry precisely so that it can isolate the steel. Do not read these numbers as a prediction of how a specific knife will feel.
We do not publish prices here. Steel prices move constantly, vary by supplier and by form, and no manufacturer publishes them in a citable way. Rather than print a number we cannot source, we have left it out and told you why. Our knife steel buying guide covers the practical price tiers in prose.
Frequently asked questions
What is the best knife steel?
There isn’t one, and the measured data is the clearest proof of that: toughness and edge retention are negatively rank-correlated at −0.70, so anything that wins on one axis is very likely losing on the other. What the data does support is a set of conditional answers. For a single knife that has to do everything, CPM MagnaCut and CPM CruWear are the two steels that refuse the usual trade-off. For a fixed blade you intend to abuse, CPM 3V at 41.5 ft-lb is not close to being beaten. For cutting abrasive material all day, CPM S90V at 779 mm and CPM S110V at 738 mm are the measured leaders. For a knife you sharpen yourself on a Sunday, AEB-L and 14C28N. Pick the use case in the tool above and it will do this arithmetic for you rather than asserting it.
Is MagnaCut really as good as everyone says?
On the measured axes, largely yes, and for a specific reason. It posts 17.6 ft-lb of toughness at 60.4 HRC with a 9.5 out of 10 corrosion rating. No other steel in this set holds a corrosion rating above 9 while staying in double-figure toughness — M390 manages 9.0 corrosion at 4.5 ft-lb, and CPM S110V manages 9.0 at 4.0 ft-lb. The honest caveat is that its CATRA edge-retention number has not been published in the same series as the others, so the tool will not rank it, and 14C28N still out-toughs it at 30.3 ft-lb. MagnaCut’s claim is not that it wins any single axis. It is that it refuses to lose badly on any of them.
S30V or S35VN?
S35VN, unless you cut abrasive material for a living. Measured: S35VN is 73 percent tougher (9.5 versus 5.5 ft-lb) at essentially the same hardness, and gives up 13 percent of edge retention (520 versus 596 mm CATRA). Corrosion resistance is identical at 7.5 out of 10. The usual description — “S35VN is a bit tougher and a bit easier to machine” — badly undersells a 73 percent difference in the property that determines whether your edge chips.
Why does cheap 420HC beat expensive steels on toughness?
Because toughness and wear resistance come from opposite places in the metallurgy. Carbides — hard particles of vanadium, chromium or niobium carbide — are what let a steel keep cutting through abrasive material, and they are also crack initiation sites. 420HC has very few of them and runs soft, at 57.5 HRC, so it measures 40.8 ft-lb. M390 is full of them and runs at 61.9 HRC, so it measures 4.5. Neither result is a defect; they are different tools. What is misleading is a chart that presents them as points on one quality ladder.
What does the CATRA number actually mean?
CATRA is a machine that pushes a blade through stacked card impregnated with silica, and reports total cards cut, expressed here in millimeters of card. It is the closest thing knives have to a standardized cutting test, and its great virtue is that it fixes edge angle and finish so the steel is the only variable. Its limitation is the same thing: real cutting is not silica card at 15 degrees per side, and the ranking it produces is a ranking of resistance to abrasive wear specifically, not of “sharpness” or of how a knife feels in the hand.
Why do some steels have blank cells?
Because the test was never published for them. Six steels here have no published Charpy toughness figure — 8Cr13MoV, AUS-8, CPM 4V, K390, Maxamet and CPM Rex 45 — and seven have no published CATRA figure. Larrin Thomas has said in print that he did not test AUS-8 or 8Cr13MoV. We could fill those cells by interpolating from chemistry or by copying a similar steel’s number, and every dot chart effectively does. We would rather show you the hole.
Does the heat treatment really change the answer?
Enough to reorder the chart. In Larrin Thomas’s own austenitising series for Nitro-V, the same steel measured 13.8 ft-lb at 59.9 HRC, 21.3 at 61.2, and 17.7 at 62.2 depending only on the temperature it was hardened from. That is a swing of more than 50 percent within one steel, from one variable, and it is larger than the gap between many of the steels on this chart. Whenever a steel comparison and your real-world experience disagree, heat treatment is the first place to look.
Can I reuse this data?
Yes. The CSV button downloads all 25 steels with every measured value, the hardness each was measured at, the chemistry and the caveats, published under a Creative Commons Attribution 4.0 licence. The underlying measurements are Larrin Thomas’s work at KnifeSteelNerds and the datasheets are the manufacturers’ — both are linked in the sources below, and if you are going to build on this, cite them as well as us.
Sources
Every toughness, edge-retention and corrosion figure on this page comes from published testing by metallurgist Larrin Thomas at KnifeSteelNerds. Every chemistry figure comes from a manufacturer datasheet, except 8Cr13MoV and AUS-8, where no maker publishes one and a compiled nominal composition is used and labeled as such. We measured nothing ourselves; we collected, cross-checked, and refused to fill the gaps.
- https://knifesteelnerds.com/2021/10/19/knife-steels-rated-by-a-metallurgist-to… — Larrin Thomas’s master ratings article. Source of the 0-10 corrosion resistance ratings for every steel (three ratings tables: stainless, high-alloy/h
- https://knifesteelnerds.com/wp-content/uploads/2021/10/stainless-ratings-table… — Readable image table of KSN 0-10 toughness / edge retention / corrosion ratings for all stainless steels (14C28N, 154CM, 420HC, AEB-L, AUS-8/8Cr13MoV,
- https://knifesteelnerds.com/wp-content/uploads/2021/10/high-alloy-steel-rating… — Readable image table of KSN 0-10 ratings for high-alloy and high-speed steels (3V, CPM-CruWear, D2, K390, Maxamet, Rex 45/HAP40, V4E/4V).
- https://knifesteelnerds.com/wp-content/uploads/2021/10/low-alloy-steel-ratings… — Readable image table of KSN 0-10 ratings for carbon/low-alloy steels, including 1095 (toughness 4.5, edge retention 1.5, corrosion 0).
- https://knifesteelnerds.com/2020/05/01/testing-the-edge-retention-of-48-knife-… — Larrin’s primary CATRA study of 48 steels, including the full heat-treatment-and-results table image.
- https://knifesteelnerds.com/wp-content/uploads/2020/04/CATRA-heat-treatments-c… — THE key numeric CATRA table: steel, austenitize temp, soak time, quench, cryo, temper, measured HRC, and TCC in mm. Source of the exact TCC values for
- https://knifesteelnerds.com/2021/05/10/edge-retention-testing-of-seven-more-st… — Follow-up CATRA study covering Maxamet, CPM Rex 45 and 420. Confirms these were tested only at high hardness (Maxamet 68.6 HRC, Rex 45 66.5 HRC, 420 5
- https://knifesteelnerds.com/2020/06/15/what-is-the-best-budget-knife-steel/ — Larrin’s explicit statement that he did NOT test AUS-6, AUS-8 or 8Cr13MoV — the basis for marking those steels’ toughness and CATRA [VERIFY]. Also the
- https://knifesteelnerds.com/wp-content/uploads/2019/12/corrosion-rating-table-… — KSN’s calculated/predicted 0-10 corrosion rating table with the Cr, Mo, W and N in solution at a stated austenitizing temperature for ~45 steels, incl
- https://knifesteelnerds.com/2019/03/04/all-about-aeb-l/ — AEB-L history (developed 1960s by Uddeholm; the 1928 patent is actually AEB/AEB-H), 14C28N history (developed by Sandvik for Kershaw, knives sold by 2
- https://knifesteelnerds.com/wp-content/uploads/2019/03/AEB-L-toughness-table2…. — Tabular AEB-L toughness: prequench, austenitize, cryo, temper, hardness and ft-lbs for ten heat treatments (23.3 ft-lbs @ 61.6 HRC standard; 37.6 @ 60
- https://knifesteelnerds.com/2019/09/23/nitro-v-its-properties-and-how-to-heat-… — Nitro-V composition table supplied by Peter Bruno of New Jersey Steel Baron, its 2017 release date, its origin as a Buderus-produced nitrogen+vanadium
- https://knifesteelnerds.com/2019/12/16/vg10-and-super-gold-2-takefu-stainless-… — VG-10 toughness stated numerically in text (5.8 ft-lbs at ~60.7 HRC, 1975F/LN2/400F) and the ~1959 development date reasoning from Takefu’s own websit
- https://knifesteelnerds.com/2018/11/05/all-about-d2-steel-knives/ — D2 patent history (filed 30 June 1927 by Gregory Comstock of Firth-Sterling; US Patent 1,695,916 issued 18 December 1928), CPM-D2 released ~2007, and
- https://knifesteelnerds.com/2018/06/04/toughness-testing-cru-wear-z-wear/ — CPM CruWear / Z-Wear toughness data table with individual specimen values, averages, standard deviations and hardness (20.1 ft-lbs @ 61.7 HRC standard
- https://knifesteelnerds.com/2020/09/28/s90v-and-s125v-knife-steel-history-prop… — CPM S90V history (originally CPM-420V, released ~1996), Bohler M390 patent 1988, Uddeholm Elmax released ~1990.
- https://knifesteelnerds.com/2020/11/23/cpm-s110v-steel-history-and-properties/ — S110V history (patent filed 2006, original datasheet September 2008, revised composition July 2010) and the labeled toughness comparison chart giving
- https://knifesteelnerds.com/2020/06/01/m390-steel-history-and-properties-and-2… — M390 history (patent application 1989, derived from K190/D7), the 20CV and CTS-204P commercial timeline (Timken-Latrobe 2005; Carpenter Micromelt 20-4
- https://knifesteelnerds.com/wp-content/uploads/2020/05/M390-toughness-comparis… — M390 toughness chart: 4.55 ft-lbs @ ~61.9 HRC (2100F/500F, recommended) vs 8.2 and 8.8 @ ~61.8 HRC (2140F/365F, retracted as retained-austenite-inflat
- https://knifesteelnerds.com/2019/11/01/crucible-s45vn-steel/ — S45VN launch article (November 2019) and the dedicated S45VN toughness chart giving 6.2 ft-lbs @ 61.4 HRC at 2000F.
- https://knifesteelnerds.com/2020/01/13/s35vn-steel-properties-and-how-to-heat-… — S35VN released in 2009 by Crucible, and the zoomed stainless toughness chart used to resolve S30V, S35VN, S45VN, Elmax, VG10, 154CM and M390 points.
- https://knifesteelnerds.com/2020/03/16/how-to-heat-treat-cpm-154-toughness-exp… — The definitive CPM-154 vs 154CM toughness chart with every point labeled by heat treatment (CPM-154 8.45 ft-lbs @ 61.6 HRC; 154CM 5.4 @ 60.2), plus L
- https://knifesteelnerds.com/2018/05/21/154cm-development-properties-legacy/ — 154CM developed by Crucible in 1959 with the Wright Air Development Center; Bob Loveless began using it in 1972; CPM-154 first produced 2004 with broa
- https://knifesteelnerds.com/2025/04/15/cpm-3v-still-the-best-high-toughness-st… — CPM 3V patented 1997 and released around 1999, plus the 2025 retest chart (41.5 ft-lbs longitudinal / 38.7 transverse at ~61.5 HRC on new material; ol
- https://knifesteelnerds.com/2021/03/25/cpm-magnacut/ — MagnaCut announcement (25 March 2021), recommended heat treat and hardness (61-62.5 HRC at 2050F/350F), corrosion testing versus 20CV and Vanax, and t
- https://nsm-ny.com/content/data-sheets/cpm-magnacut-data-sheet.pdf — Crucible CPM MagnaCut datasheet: composition (C 1.15, Cr 10.70, V 4.00, Mo 2.00, Nb 2.00, N 0.20), aim hardness 60-63 HRC, carbide type/volume, and Cr
- https://nsm-ny.com/content/data-sheets/cpm-s30v-data-sheet.pdf — Crucible CPM S30V composition (C 1.45, Cr 14.00, Mo 2.00, V 4.00), aim hardness 58-61 HRC, Charpy C-notch 10.0 ft-lbs, CATRA 145% of 440C ESR.
- https://nsm-ny.com/content/data-sheets/cpm-s35vn-data-sheet.pdf — Crucible CPM S35VN composition (C 1.40, Nb 0.50, Cr 14.00, Mo 2.00, V 3.00) and aim hardness 58-61 HRC.
- https://nsm-ny.com/content/data-sheets/cpm-s45vn-data-sheet.pdf — Crucible CPM S45VN composition (C 1.48, Cr 16.00, V 3.00, Mo 2.00, Nb 0.50, N 0.15), aim hardness 59-61 HRC, carbide type/volume.
- https://nsm-ny.com/content/data-sheets/cpm-s90v-data-sheet.pdf — Crucible CPM S90V composition (C 2.30, Cr 14.00, Mo 1.00, V 9.00) and aim hardness HRC 56/59.
- https://nsm-ny.com/content/data-sheets/cpm-s110v-data-sheet.pdf — Crucible CPM S110V composition (C 2.80, Nb 3.00, Co 2.50, Cr 15.25, Mo 2.25, V 9.00) and typical application hardness 58/61 HRC.
- https://nsm-ny.com/content/data-sheets/cpm-cruwear-data-sheet.pdf — Crucible CPM CRUWEAR composition (C 1.15, Cr 7.50, W 1.00, Mo 1.60, V 2.40), heat treatable to HRC 60-65, aim hardness HRC 62.
- https://nsm-ny.com/content/data-sheets/cpm-3v-data-sheet.pdf — Crucible CPM 3V composition (C 0.80, Mn 0.30, Si 1.00, Cr 7.50, Mo 1.30, V 2.75) and aim hardness HRC 58-60.
- https://nsm-ny.com/content/data-sheets/cpm-4v-data-sheet.pdf — Crucible CPM 4V composition (C 1.35, Mn 0.40, Si 0.80, Cr 5.00, Mo 2.95, V 3.85) and aim hardness HRC 62-64.
How to use this responsibly. These are laboratory measurements on specimens heat treated to a metallurgist’s specification, not predictions about any particular knife. A production blade in the same steel may perform very differently depending on its heat treatment, its geometry, its thickness behind the edge and how it is sharpened. Nothing here is a safety rating: no steel on this chart is unbreakable, and prying with any knife is a good way to find that out. Popular EDC has no commercial relationship with any steel manufacturer named here. Data last verified August 2026.
In short: M390 holds an edge slightly longer than CPM MagnaCut. MagnaCut is far tougher and more corrosion resistant, which decides most real choices.
The comparisons people actually run
These are the head-to-heads that come up most often. Each verdict is read off the same published ratings the tool uses — toughness, edge retention and corrosion resistance on a 0 to 10 scale — and each link opens the comparison pre-loaded so you can see the underlying numbers and chemistry side by side.
M390 vs MagnaCut
M390 holds an edge slightly longer. MagnaCut is dramatically tougher — 7 against 3.5 — and a shade more corrosion resistant. If you are choosing a folder for hard use, MagnaCut is the safer blade; if you cut abrasive material all day and treat the knife gently, M390 keeps its edge longer.
M390 / 20CV / 204P — toughness 3.5, edge retention 6.5, corrosion 9. CPM MagnaCut — toughness 7, edge retention 5, corrosion 9.5.
Open this comparison in the tool
14C28N vs D2
Not close, except on edge retention. D2 holds an edge longer (5 against 3), but 14C28N is rated 9 for toughness against D2's 3.5 and 8.5 for corrosion resistance against 4.5. D2 is semi-stainless and will spot if you ignore it. For most carriers 14C28N is the better everyday steel and it usually costs less.
14C28N — toughness 9, edge retention 3, corrosion 8.5. D2 — toughness 3.5, edge retention 5, corrosion 4.5.
Open this comparison in the tool
S45VN vs MagnaCut
MagnaCut wins on the two columns that separate them: toughness 7 against 4, corrosion 9.5 against 8. S45VN edges it very slightly on edge retention. This is the clearest case on the whole chart of a newer steel simply being better, which is why Chris Reeve ships both and prices MagnaCut above it.
S45VN — toughness 4, edge retention 5.5, corrosion 8. CPM MagnaCut — toughness 7, edge retention 5, corrosion 9.5.
Open this comparison in the tool
S30V vs S35VN
S35VN was designed as the fix for S30V's brittleness and the ratings show it: toughness 5 against 4, with identical corrosion resistance. S30V is rated slightly higher on edge retention. In practice S35VN is the more forgiving blade and S30V is now the cheaper one.
S30V — toughness 4, edge retention 6, corrosion 7.5. S35VN — toughness 5, edge retention 5, corrosion 7.5.
Open this comparison in the tool
S35VN vs S45VN
S45VN buys corrosion resistance (8 against 7.5) and a little edge retention; S35VN keeps a toughness advantage (5 against 4). They are close enough that heat treatment and price should decide it.
S35VN — toughness 5, edge retention 5, corrosion 7.5. S45VN — toughness 4, edge retention 5.5, corrosion 8.
Open this comparison in the tool
CPM 3V vs MagnaCut
3V is the toughest steel in common production, rated 9, and MagnaCut is close behind at 7. The real gap is rust: MagnaCut is rated 9.5 for corrosion resistance against 3V's 5.5. For a fixed blade that lives dry, 3V. For anything that meets sweat or salt water, MagnaCut.
CPM 3V — toughness 9, edge retention 4.5, corrosion 5.5. CPM MagnaCut — toughness 7, edge retention 5, corrosion 9.5.
Open this comparison in the tool
MagnaCut vs S35VN
Identical edge retention, with MagnaCut ahead on toughness (7 against 5) and corrosion resistance (9.5 against 7.5). Crucible's own datasheets put CATRA edge retention at 135 percent against 130 percent of a 440C reference, and Charpy toughness at 38 ft-lbs against 25.
CPM MagnaCut — toughness 7, edge retention 5, corrosion 9.5. S35VN — toughness 5, edge retention 5, corrosion 7.5.
Open this comparison in the tool
Ratings are Larrin Thomas's published figures, cited in full in the sources above. Heat treatment can move any of these outcomes — a well-treated blade in the "losing" steel will beat a poorly treated one in the winner.
How this was put together. Every figure on this page is traced to a manufacturer datasheet or standards document, listed in the sources above. Written and maintained by James Nicholas.
