AAV / Gene Therapy CDMOs — Ranked by Signal Score · 2026

· updated as new inspection, certificate, and trial data enters the platform

AAV / Gene Therapy manufacturers on CDMO Signal: 52 listed. 7 are fully evidenced (all four Signal Score pillars and at least one ClinicalTrials.gov match). The highest Signal Scores in that fully evidenced set are Takeda (also makes its own products, Signal Score 83.4), AGC Biologics (Signal Score 79.9) and Thermo Fisher Scientific (Signal Score 77.6). Outside that set, higher Signal Scores on this list include Pharmaron (Signal Score 86.9), FUJIFILM Diosynth Biotechnologies (Signal Score 86.7) and Charles River Laboratories (Signal Score 86.3). A higher score there is not a fully evidenced record. The score does not measure serotype experience, suspension versus adherent scale, or full versus empty capsid analytics. Signal Score is computed from FDA inspection records, EMA EudraGMDP and MHRA GMP certificates, and ClinicalTrials.gov. CDMOs cannot pay to change it. Updated 2026-11-20. Methodology.

AAV CDMOs with the most regulatory evidence on file

Catalent
76.1
Tampa, FL
24 FDA inspections · 39 EMA/MHRA GMP certs
Last FDA inspection: 2026-05-22 · VAI
Warning letter 2025-11-20
See the evidence →
FUJIFILM Diosynth Biotechnologies
86.7
Research Triangle, NC · Holly Springs, NC · Hillerod, DK
0 FDA inspections · 35 EMA/MHRA GMP certs
Last FDA inspection: None on file
See the evidence →
AGC Biologics
79.9
Longmont, CO · Milan, IT · Copenhagen, DK · Chiba, JP
4 FDA inspections · 27 EMA/MHRA GMP certs
Last FDA inspection: 2026-05-29 · NAI
See the evidence →
Thermo Fisher Scientific
77.6
Plainville, MA · Alachua, FL · Cambridge, MA
19 FDA inspections · 10 EMA/MHRA GMP certs
Last FDA inspection: 2026-06-30 · NAI
See the evidence →
Takeda
83.4
Boston, MA · Zurich, CH
4 FDA inspections · 20 EMA/MHRA GMP certs
Last FDA inspection: 2026-04-10 · NAI
See the evidence →
52
CDMOs tracked
52
Fully scored
17
With FDA records
96
FDA inspections
219
EMA/MHRA GMP certs
2
Warning letters

Shortlist AAV / gene therapy CDMOs with Signal

CDMO Signal is an independent sourcing platform for comparing and shortlisting AAV and gene therapy CDMOs before an RFI. This page is the entry point — the ranked table below is derived from FDA inspection history, EMA/MHRA GMP certificate density, ClinicalTrials.gov sponsor and program footprint, and monitored capacity signals. Use it to narrow the field, then evaluate whether each shortlisted supplier's public evidence supports your serotype, scale, and analytics requirements before contacting them.

  1. Compare AAV and gene therapy CDMOs side-by-side on FDA inspection history, EMA/MHRA GMP certificate density, ClinicalTrials.gov sponsor and program footprint, and monitored capacity signals — every number traceable to primary sources.
  2. Filter by regulatory jurisdiction, clean-inspection posture, and clinical-program breadth to your program's shortlisting criteria.
  3. Shortlist Use the ranked results to build your shortlist, then open each profile to review the underlying evidence before contacting suppliers. Follow up with the shortlisted CDMOs to evaluate whether their public evidence — and their direct answers to your RFI — support your serotype, scale, and analytics requirements.

Adeno-associated virus (AAV) is the dominant delivery platform for in vivo gene therapy, from rare-disease programs to larger systemic indications. Rising demand has pushed CDMOs to scale suspension processes and tighten full/empty capsid analytics.

Selecting an AAV CDMO hinges on serotype experience, suspension vs adherent platform, full/empty capsid separation, and a proven regulatory record. The rankings below score each manufacturer on FDA inspections, GMP certification, clinical activity, and capacity.

FDA inspection outcomes across these CDMOs: 36 NAI (no action), 58 VAI (voluntary action), 2 OAI (official action). Leading inspection & GMP sites: United States (72), UNITED KINGDOM (55), Denmark (52), Germany (24), Italy (15).

Last updated 2026-11-20. Sourced from FDA, EMA EudraGMDP, MHRA GMDP, and ClinicalTrials.gov.

Ranked by Signal Score (0–100). Fully evidenced marks a CDMO scored on all four pillars with at least one confirmed clinical program. Partly scored and profile-only CDMOs follow the scored rows. How we score

Showing all operator types.

# CDMO Signal Score Quality FDA · GMP Last FDA inspection Capacity Programs
1
PB
San Diego, CA · Shaoxing, China, Liverpool, UK, Ningbo China
86.9 100.0 5 insp · 1 GMP 2026-08-14 · NAI Available —
2
FD
Research Triangle, NC · Holly Springs, NC · Hillerod, DK
Hiring · 20 open roles
86.7 100.0 0 insp · 35 GMP None on file Expanding —
3
CR
Newbury Park, CA · Memphis, TN · Keele, UK · Rockville, MD
86.3 99.2 9 insp · 11 GMP 2026-05-01 · NAI Limited —
4
KB
Durham, NC · Boulder, CO
Hiring · 97+ open roles
84.1 98.5 2 insp · 1 GMP 2026-04-01 · VAI Limited —
5
AJ
San Diego, CA · Osaka, JP
83.5 100.0 0 insp · 1 GMP None on file — —
6
TK
Boston, MA · Zurich, CH
Fully evidenced
Also makes its own products
83.4 100.0 4 insp · 20 GMP 2026-04-10 · NAI — 100 programs
7
RP
Keele, UK · Sodertalje, SE
82.9 100.0 0 insp · 3 GMP None on file — —
8
WB2
Shanghai, CN · Wuxi, CN
82.7 99.6 1 insp · 4 GMP 2026-07-03 · VAI — —
9
LB
Incheon, KR · Syracuse, NY
Hiring · 8 open roles
81.6 100.0 2 insp · 0 GMP 2026-09-18 · NAI — —
10
NW
Vilnius, LT; Waltham (MA), USA
Hiring · 16 open roles
80.8 100.0 0 insp · 8 GMP None on file Available —
11
BR
Glasgow, UK · Rockville, MD
Hiring · 16 open roles
Also makes its own products
80.3 97.2 6 insp · 10 GMP 2026-05-13 · VAI — —
12
AG
Longmont, CO · Milan, IT · Copenhagen, DK · Chiba, JP
Hiring · 38 open roles
Fully evidenced
79.9 99.8 4 insp · 27 GMP 2026-05-29 · NAI Available 2 programs
13
SK
Dublin, Ireland. Rancho Cordova, CA. King of Prussia, PA. Houston, TX. Korea. France.
79.5 100.0 — None on file Available —
14
BV
Turku, Finland
Hiring · 2 open roles
79.4 99.8 1 insp · 4 GMP 2024-12-19 · VAI — —
15
UM
Boston, MA
78.7 96.7 2 insp · 0 GMP 2025-12-18 · VAI — —
16
GZ
Indianapolis, IN · Lexington, MA
Hiring · 26 open roles
78.1 100.0 1 insp · 0 GMP 2025-08-27 · NAI Available —
17
AA
Rockville, MD · Alachua, FL · Munich, DE
78.1 100.0 0 insp · 1 GMP None on file — —
18
NG
Leiden, NL
78.1 100.0 0 insp · 1 GMP None on file — —
19
ET
Jumet, BE
78.1 100.0 0 insp · 4 GMP None on file — —
20
PG
Berlin, DE
78.1 100.0 0 insp · 4 GMP None on file — —
21
CE
Basel, Switzerland
Hiring · 4 open roles
78.1 100.0 0 insp · 1 GMP None on file — —
22
XP
Bristol, United Kingdom
78.1 100.0 0 insp · 1 GMP None on file — —
23
AD
London, United Kingdom
78.1 100.0 0 insp · 2 GMP None on file — —
24
TF
Plainville, MA · Alachua, FL · Cambridge, MA
Hiring · 299 open roles
Fully evidenced
77.6 93.2 19 insp · 10 GMP 2026-06-30 · NAI Available 2 programs
25
AG2
San Carlos, CA
77.6 98.8 1 insp · 0 GMP 2025-04-18 · VAI — —
26
VG
San Sebastián, Spain
Hiring · 3 open roles
76.8 100.0 0 insp · 4 GMP None on file Available —
27
OX
Oxford, United Kingdom
Hiring · 4 open roles
Fully evidenced
Also makes its own products
76.2 100.0 2 insp · 13 GMP 2024-06-17 · NAI Available 7 programs
28
Tampa, FL
Hiring · 311 open roles
76.1 77.3 24 insp · 39 GMP 2026-05-22 · VAI
Warning letter 2025-11-20
— —
29
CI
Cincinnati, OH
Hiring · 2 open roles
Fully evidenced
Academic / national centre
75.1 99.6 2 insp · 0 GMP 2025-02-05 · NAI — 1 program
30
DH
Washington, DC
Platform provider
73.0 — — None on file — —
31
Basel, Switzerland
Hiring · 100+ open roles
Fully evidenced
72.8 86.6 11 insp · 8 GMP 2025-09-29 · VAI — 1 program
32
RT
Rome, IT
Fully evidenced
71.1 100.0 0 insp · 4 GMP None on file — 1 program
33
IU
Indianapolis, IN
Academic / national centre
66.3 — — None on file — 49 programs
34
EM
Baltimore, MD
Also makes its own products
66.1 — — None on file — 4 programs
35
CG
Marlborough, MA · Uppsala, SE
Platform provider
65.5 — — None on file — —
36
VE
Rockville, MD · Memphis, TN · Keele, United Kingdom
64.0 — — None on file — —
37
AN
Columbus, OH
63.0 — — None on file Available —
38
RA
Stockholm, Sweden
63.0 — — None on file — —
39
AB
Research Triangle, NC · San Sebastian, ES
Also makes its own products
61.8 — — None on file — 10 programs
40
OB
Hamilton, ON
61.5 — 0 insp · 2 GMP None on file — —
41
FB
Columbus, OH
Hiring · 15 open roles
61.2 — — None on file Expanding 3 programs
42
MA
College Station, TX
59.0 — — None on file Available —
43
VB
Philadelphia, PA
59.0 — — None on file — —
44
MS
Princeton, NJ
59.0 — — None on file — —
45
WB
59.0 — — None on file — —
46
EB
Waltham, MA
Hiring · 8 open roles
59.0 — — None on file — —
47
CP
Stevenage, United Kingdom
Academic / national centre
59.0 — — None on file — —
48
VM
Cambridge, MA
59.0 — — None on file — —
49
AM
Gerenzano, Italy
59.0 — — None on file — —
50
HX
Seoul, KR
Also makes its own products
59.0 — — None on file — —
51
UB
Shanghai, CN · Boston, MA
Hiring · 1 open role
56.5 — — None on file — —
52
CV
Cologne, Germany
56.5 — — None on file — —
How we score CDMOs →

What to evaluate in a AAV / Gene Therapy CDMO

Platform & scale
Suspension processes scale better than adherent (HEK293) for commercial volumes. Confirm the CDMO's largest validated bioreactor scale and serotype track record.
Capsid quality
Full/empty capsid ratio and aggregate control drive potency and safety. Ask about AUC, anion-exchange separation, and the analytics package for capsid characterization.
Regulatory readiness
For programs heading to BLA, a strong FDA inspection history and EMA/MHRA GMP coverage matter — both are shown above and on each profile.
Capacity & timelines
AAV manufacturing slots can be constrained. Weigh the capacity signal and clinical-program load before committing.

Suspension vs adherent AAV — which platforms do CDMOs use?

AAV manufacturing uses two upstream platforms: adherent (HEK293 on flasks or fixed-bed bioreactors) and suspension (HEK293-based cells in stirred-tank bioreactors). Suspension scales more cleanly for commercial volumes and is the industry direction, but many CDMOs still support adherent for smaller-batch and rare-disease programs. Larger diversified viral vector CDMOs typically run both. The specific platform a CDMO uses for a given serotype is an important early diligence question; ask about validated bioreactor scale for the target serotype and the productivity (vg/L) achieved.

Full/empty capsid capabilities — how AAV CDMOs differ

Full/empty capsid ratio is the AAV analytics variable that most drives potency and safety. CDMOs differ in the maturity of their downstream separation — analytical ultracentrifugation (AUC), anion-exchange chromatography, cesium chloride gradient — and in operator experience with each technique. Signal Score does not yet track per-CDMO capsid separation platforms directly, so sponsors should raise the question during vendor selection. Individual CDMO profile pages capture press-monitored capacity expansions that often flag downstream investment; the methodology page outlines what is and isn't in scope.

Which AAV CDMOs handle commercial-scale supply?

Commercial-scale AAV manufacturing is a rapidly maturing segment. The commercial-scale AAV filter surfaces CDMOs with top-quartile Signal Score or broad ClinicalTrials.gov footprint. Commercial AAV has historically been dominated by pure-play viral vector specialists, but larger diversified CDMOs have added capacity aggressively over the last three years. Sponsors weighing commercial supply should verify not just capacity headroom but demonstrated tech transfer of full/empty capsid analytics and consistency across scale-up runs — this is where late-stage programs most often stumble.

How CDMO Signal computes the AAV ranking

The AAV Signal Score is the platform's standard four-pillar composite. Quality Compliance (35%) leans on FDA inspection classifications (NAI, VAI, OAI), Warning Letter history, and EMA/MHRA GMP certificate density. Operations (25%) reflects ClinicalTrials.gov sponsor and program footprint matched to each CDMO. Financial Stability (20%) draws on public/private status and parent-company backing. Capacity Intelligence (20%) counts the manufacturing sites and modalities listed on the profile; job postings are not a score input. Press announcements and SEC filings are shown on profiles as company signals and do not change any score. For AAV specifically, Quality Compliance and Capacity Intelligence disproportionately signal — this is a modality where facility investment and regulatory posture directly enable or block commercial supply. Full weightings at cdmosignal.com/methodology.

Serotype experience — AAV2, AAV5, AAV6, AAV8, AAV9 across CDMOs

AAV serotypes have distinct tropism and manufacturing characteristics. AAV2 is the historical reference serotype with the deepest CDMO experience base but modest yields. AAV5 and AAV8 are common for liver and neuromuscular indications. AAV9 crosses the blood-brain barrier and dominates CNS gene therapy programs but presents higher immunogenicity considerations. AAV6 is used in ex vivo genome editing workflows. Most CDMOs specialize in a subset — pure-play viral vector CDMOs typically have qualified processes across 3-5 serotypes; larger diversified CDMOs may cover 6+. Sponsors should ask for validated productivity (vg/L or vg/cell) at the target serotype specifically, since yields vary 3-5× across serotypes on the same platform. The individual CDMO profile pages capture qualified serotypes where disclosed.

Which AAV CDMOs manufacture FDA-approved commercial products?

Commercial AAV manufacturing is a very narrow field. Approved AAV gene therapies to date include Luxturna (Spark, AAV2), Zolgensma (Novartis, AAV9), Hemgenix (CSL, AAV5), Roctavian (BioMarin, AAV5), Elevidys (Sarepta, AAVrh74), and Casgevy (uses CRISPR-Cas9 with AAV6 for delivery). Manufacturing has historically been done at the innovator or through a small set of specialized CDMOs including Oxford Biomedica, Catalent, Thermo Fisher, and Charles River. This is the highest-bar quality signal — sponsors targeting commercial supply should treat commercial AAV track record as a hard qualifying criterion, and verify not just "has capacity" but "has delivered validated commercial lots." The commercial-scale AAV filter surfaces CDMOs with the strongest current signal.

Rare disease AAV — batch size, dose, and cost considerations

Rare disease AAV programs face a distinctive economics: extremely high per-patient dose (often 10^13-10^14 vg/kg for systemic delivery, 10^15 vg total for a 100kg adult) combined with small patient counts (hundreds to low thousands globally). This flips the standard commercial-scale calculus — you need enough capacity for high-dose batches but may only run a few campaigns per year. Sponsors should verify the CDMO's largest validated batch is sized appropriately (typically 500L-2000L suspension for systemic dosing), and separately confirm their small-batch adherent option remains available for lower-dose pediatric or CNS-injected programs. Cost per dose at rare disease scale often runs six figures at the CDMO level alone; that pricing dynamic is what makes tech transfer risk especially costly to correct mid-program.

How to structure an AAV CDMO RFP — a 6-step buyer framework

For sponsors running competitive AAV CDMO selection: 1. Serotype-specific yield disclosure — ask each CDMO for their validated productivity in vg/cell or vg/L for your target serotype, not their platform average. 2. Full/empty separation approach — AUC, anion-exchange, or CsCl, and demonstrated DAR consistency across scale-up runs. 3. Analytics package — potency assays (typically transduction-based cell assays), residual DNA specs, capsid identity confirmation. 4. Regulatory posture — FDA inspection classifications specifically on the facility slated for your program (see ranked table above and each CDMO profile). 5. Timeline realism — first-in-human material typically 12-18 months from tech transfer; commercial supply 3-5 year engagement with tech-transfer costs weighted heavily to year one. 6. Capacity headroom — verify the CDMO isn't slot-constrained by other programs in your target campaign window. Comparability data between the CDMO's development runs and their validated GMP process is what most sponsors underestimate.

AAV / Gene Therapy CDMOs — Frequently Asked Questions

Who are the top AAV / Gene Therapy CDMOs?
Among AAV / Gene Therapy manufacturers with a fully evidenced Signal Score (all four pillars and at least one ClinicalTrials.gov match), the highest are Takeda, AGC Biologics, Thermo Fisher Scientific. The table above is the full list. CDMOs cannot pay to change a score.
How many AAV / Gene Therapy CDMOs have FDA inspection records?
CDMO Signal tracks 17 AAV / Gene Therapy CDMOs with FDA inspection records and 219 EMA/MHRA GMP certificates across the group.
What should I look for in an AAV CDMO?
Look for serotype experience, a scalable suspension platform, strong full/empty capsid analytics, and a clean FDA and EMA/MHRA record. Match validated bioreactor scale to your dose and indication.
What's the difference between AAV and lentiviral CDMOs?
AAV is non-integrating and produced at higher volumes for in vivo gene therapy. Lentivirus integrates into the genome, is made under BSL-2 with RCL testing, and is used mainly for ex vivo cell engineering like CAR-T.
Compare side-by-side
See two CDMOs' scoring, inspection history, and clinical footprint next to each other.
Open comparison →
Run a sourcing search
Describe your program's requirements — Signal matches CDMOs against FDA, EMA/MHRA, ClinicalTrials.gov and capacity signals.
Search for a CDMO →
Save this AAV shortlist
Get an email when a CDMO on this list has a new FDA inspection, GMP certificate, or scoring change.
Related modalities
Lentiviral Vector CDMOs Gene Editing CDMOs Plasmid DNA CDMOs
Other Modalities
CAR-T Lentiviral Cell Therapy mRNA/LNP Plasmid DNA Biologics Oligo/ASO Adenoviral Gene Editing Exosome Recombinant Proteins ADC Gene Therapy Cell & Gene Therapy Viral Vector