Prefixes from A to M cover PRESTIGE (ABF, ABW, ABP), Attackers (ADN,
ATID), kawaii* (CAWD), Das! (DASS), E-BODY (EBWH), FALENO (FSDSS),
IdeaPocket (IPZZ, IPX), Madonna (JUQ, JUL), Fitch (JUFE), MOODYZ
(MIDV, MIAA) and Tameike Goro (MEYD), among others. The full table is
below in alphabetical order.
Prefixes N through Z are in the second half of this lookup. If you would rather browse by label than by letter, the studio code reference groups the same prefixes by the company that issues them.
What do the columns mean?
Prefix is the letter portion of the code, Label is the issuing company or line, and Status says whether the prefix still receives new numbers.
- Current — actively receiving new serial numbers
- Retired — back catalogue intact, no new numbers issued
- Blank — we have not confirmed the status
Retired never means unusable. Every code ever issued under a retired prefix still identifies its release permanently. It only means that a serial higher than the last one issued does not exist, which is worth knowing before you spend time hunting for one.
Which label owns each prefix from A to M?
| Prefix | Label | Status |
|---|---|---|
ABF |
PRESTIGE | Current |
ABP |
PRESTIGE | Retired |
ABW |
PRESTIGE | Retired |
ADN |
Attackers | Current |
ATID |
Attackers | Current |
BAZX |
BAZOOKA | |
BLK |
kira☆kira | Current |
CAWB |
kawaii* | Current |
CAWD |
kawaii* | Retired |
CHN |
PRESTIGE | |
DASD |
Das! | Retired |
DASS |
Das! | Retired |
DOCP |
PRESTIGE | |
DSOD |
Das! | Current |
DV |
Alice Japan | Retired |
DVAJ |
Alice Japan | Current |
DVDES |
Deep's | Retired |
DVDMS |
Deep's | Retired |
DVMM |
Deep's | Current |
EBOD |
E-BODY | Retired |
EBWH |
E-BODY | Current |
FNS |
FALENO | Current |
FSDSS |
FALENO | Retired |
GVH |
Glory Quest | Current |
HMN |
Honnaka | Current |
HND |
Honnaka | Retired |
HUNTA |
Hunter (HHH Group) | Retired |
HUNTB |
Hunter (HHH Group) | Retired |
HUNTC |
Hunter (HHH Group) | Current |
IDBD |
IdeaPocket | Retired |
IPTD |
IdeaPocket | Retired |
IPX |
IdeaPocket | Retired |
IPZ |
IdeaPocket | Retired |
IPZZ |
IdeaPocket | Current |
JBD |
Attackers | Current |
JUFD |
Fitch | Retired |
JUFE |
Fitch | Current |
JUL |
Madonna | Retired |
JUQ |
Madonna | Retired |
JUR |
Madonna | Current |
JUX |
Madonna | Retired |
JUY |
Madonna | Retired |
KAWD |
kawaii* | Retired |
KWBD |
kawaii* | Current |
MEYD |
Tameike Goro | Retired |
MFYD |
Tameike Goro | Current |
MIAA |
MOODYZ | Retired |
MIAB |
MOODYZ | Current |
MIDA |
MOODYZ | Current |
MIDD |
MOODYZ | Retired |
MIDE |
MOODYZ | Retired |
MIDV |
MOODYZ | Retired |
MIFD |
MOODYZ | Current |
MIMK |
MOODYZ | Current |
MIZD |
MOODYZ | Current |
Attribution and status checked 2026-08-03 against each label's own release
listing and against FANZA product records. Two entries are worth calling out
because published lists get them wrong. Alice Japan has not stopped — DVAJ
was still receiving numbers in mid-2026, even though the older DV line is long
closed. And DASS reached 999 in July 2026, so Das! is now numbering DSOD
instead.
Entries we could not confirm are absent rather than guessed. Published prefix lists copy each other freely, errors included, so a plausible-looking row in a reference table is not evidence. Where we considered a prefix and could not settle it, it is recorded in the source of this page as a pending check instead of being printed as fact.
Which prefixes in this range are easiest to confuse?
Four clusters, and all four are confusions of label, not of code.
| Cluster | Prefixes | Why it confuses |
|---|---|---|
JU- |
JUR, JUQ, JUL, JUY, JUX, JUFE, JUFD |
Two different labels, split by the third letter |
IP- |
IPZZ, IPZ, IPX, IPTD |
One label, four eras; IPZ and IPZZ differ by one character |
AB- |
ABF, ABW, ABP |
One label, three eras, no visual hint of order |
MI- |
MIDA, MIDV, MIDE, MIDD, MIAB, MIAA, MIFD |
One label, several parallel and successive lines |
The general rule: within these clusters, a one-character difference indicates a different line or a different era, never a variant of the same release. If two codes differ by one letter and you assumed they were the same title in two editions, they almost certainly are not. Genuine duplicate numbering works differently — see why does the same title have more than one code.
What if your code starts with a digit or a symbol?
Then it is not a studio prefix, and this table is the wrong place to look.
| Shape | Family | Where to look |
|---|---|---|
259LUXU-1234 |
Amateur / planning series | Amateur and planning-series codes |
FC2-PPV-1234567 |
Individual-seller platform | FC2 codes vs studio codes |
010124-001 |
Uncensored subscription site | How uncensored releases are numbered |
1abf00042 |
Storefront internal ID | Strip the digits and padding: ABF-042 |
hhd800.com@ABF-042 |
Mirror site tag | Strip everything before the prefix |
The fourth row catches a surprising number of failed lookups. A leading digit followed by a recognisable prefix is a padded internal ID, not an unknown label — the mechanics are in the anatomy of a JAV code.
Why does this table not show release counts?
Because we cannot measure release counts, and a number we cannot measure is worse than no number.
You will find prefix tables elsewhere that attach a count to each row. Ours does not, for a specific reason: our index stores faces, not catalogue records. It has no maker field, no release date and no title inventory. The only quantity we could honestly derive would be how many titles bearing a prefix appear on the sites we index — which measures aggregator coverage, not studio output, and would be misread as the latter every time.
Boundary worth stating: the index spans 106 sites, of which only around a dozen carry Japanese-style codes, and 2,246 of 2,333 representative images come from a single source (our index, 2026-08 snapshot). That skew makes any per-prefix count structurally misleading.
What do you do if the prefix is right and nothing resolves?
Check whether the serial is inside the range that prefix ever reached, then switch identifiers.
A retired prefix has a last issued number. Any serial above it was never allocated, and no catalogue can produce it — that is the most common "the prefix is correct but the code fails" case, and it is covered in codes from studios that no longer release.
If the code is well-formed, in range and still returns nothing, the frame is a better handle than the number. Our index compares 512-dimension ArcFace vectors across 241,792 indexed faces and 2,333 named performers (our index, 2026-08 snapshot) at a 0.40 similarity threshold, and it does not require a code at all.