Skip to main content
INANVIL · ANVIL · PHASE 3

Design and evaluate moleculeswith manufacturability in mind.

Design siRNA / ASO / mAb / biosimilar / BsAb leads, then evaluate developability — all in one platform, with synthesis and purification in view. Avoid the 6-month CMC restart.

5 free design + 5 evaluation / month with Pro · 14-day Pro trial
Design + Evaluate · single workflowLive
DESIGN
Target genePCSK9
SpeciesHuman
Chemistry2′-OMe
Output20 cand.
Running… 30s left
EVALUATE
#CandidateScore
1siPCSK9-129.2
2siPCSK9-089.0
3siPCSK9-178.7
4siPCSK9-048.5
5siPCSK9-158.3
01THE PROBLEM

By the time you find out, you've already spent millions.

A molecule looks perfect in Discovery. Target hit. Strong affinity.

Then it enters CMC — it aggregates. It is hard to purify. PTM hot spots everywhere. Solubility at clinical dose is too low.

The team restarts 6 months later. Most of this could have been predicted from the sequence.

COST PER FAILURE STAGE
Discovery rejection$50K
Late-stage CMC failure$2-5M
Phase I withdrawal$10-20M
02THE SOLUTION

Catch it on the screen, not in the lab.

InAnvil evaluates molecules in 30 seconds across 8 developability dimensions — before you commit to a chemistry route, before you commission a CRO, before you start a tank.

03CAPABILITIES

Design + Evaluation, one platform.

11 submodules across mAb · ADC · oligo · biosimilar · BsAb — design and developability evaluation.

01
Oligo

D1 · siRNA Lead Design

Gene + species + chemistry → 10–20 ranked candidates with modifications, off-target check, synthesis cost, recommended purification.

02
Oligo

D2 · ASO Lead Design

Same plus gapmer / SSO / blocker type → ASO candidates with MOE / LNA wing and CpG screening.

03
mAb

D3 · mAb Humanization + Affinity + Liability

Murine VH/VL → humanized sequence, humanness score, liability report.

04
mAb

D4 · Biosimilar reverse design

Reference biologic + target market → CMC strategy, patent risk map, CQA alignment.

05
BsAb

D5 · Bispecific Design

Two targets + mechanism → format recommendation, full sequence, expression yield, CMC complexity.

COMING IN PHASE 4+D6 mRNA · D7 LNP · D8 TriSpec+ · D9 VHH · D10 ADC design · D11 AAV capsid
YEAR 2+Peptide · Engineered protein · CAR-T construct · Lentiviral vector · Vaccine antigen
NOT IN SCOPESmall molecules · Diagnostics · Devices · Aptamers
04UNDER THE HOOD

Built on best-in-class research tools.

We don't reinvent the wheel for prediction. We integrate the academic and industry tools that work — and calibrate them with Inscinstech's own CMC data.

Antibody design / evaluation
BioPhi (Merck/MIT)TAP / SAP (Oxford OPIG)Boltz-2 (MIT)IgFold (Apache 2.0)AbNumber / ANARCI (BSD)ProteinMPNN (MIT)
Oligonucleotide design
ViennaRNABLAST+ (Public domain)NCBI EntrezDSIR / Reynolds / Ui-Tei / Amarzguioui (re-implemented)
Inscinstech proprietary
OligoMS synthesizability scorerNestoPure purification recommendationBsAb expression yield modelCMC v2.2 calibration data
Why this matters: technical buyers (CMC heads · VP discovery) need to trust the underlying tools. Transparent disclosure = trust = lower decision friction.
05SHOWCASE · ONE-CLICK TO DEVELOPMENT

From scored to scaled.

Pass a scored molecule directly to InForge. It picks up the recommended purification route and starts process design — with your team's history and preferences pre-loaded.

INANVIL · REPORT
mAb-003 · 9.1 / 10 ★
Recommended route: Protein A → CEX → UVC inact → TFF. No major risk flags.
SEND →
INFORGE · INTAKE
mAb-003 process · v0.1 draft
Process tree pre-populated · team preferences loaded · DOE templates ready.
06USE CASES

Three teams. Three quarters saved.

LD
Lead Discovery scientist

We screen 50 candidates a year. Used to ship 5 bad ones to CMC. Now we ship 5 of the best 5.

Anonymous · v0 placeholder
VP
VP CMC

InAnvil cut our 'molecule rejected after 3 months' incidents by 70% in Year 1.

Anonymous · v0 placeholder
CR
CRO process lead

When a client sends a sequence + InAnvil report, we quote 30% faster and more accurately.

Anonymous · v0 placeholder
07PRICING

InAnvil at every tier.

Free
— not available on this tier
Pro
5 mAb scores / month
Team
50 mAb + ADC + Oligo scores · 10 Boltz-2 structures
Enterprise
Unlimited · API · private deployment
08COMPLIANCE & DEPLOY

Read-only evaluation. Your sequences stay yours.

Customer sequences never train the model
Per-tenant isolation — sequences stored in your namespace only
Private deployment available for sensitive IP
Audit log of every scoring request
CN data residency available on inscinstech.cn
09SAMPLE OUTPUT

A full developability scorecard.

app.inscinstech.ai
Candidate × Developability heatmap30s scoring
AggViscSolPTMPurFormCharImm
mAb-0018.2
mAb-0026.4
mAb-0039.1
mAb-0043.8
mAb-0058.7
AGGREGATION
0.3% HMW
RISK · LOW
VISCOSITY
12 cP @ 150 g/L
RISK · LOW
SOLUBILITY
> 200 g/L
RISK · LOW
PTM HOTSPOTS
0 critical
RISK · LOW
11FAQ

Common questions

Predictions are calibrated against Inscinstech CMC outcomes data (v2.2). Risk categories (low/med/high) are conservative; we surface uncertainty rather than over-claim precision.
For mAbs: heavy + light chain sequences (FASTA). For ADCs: antibody + linker + payload + DAR. For oligos: sequence + modifications.
Yes — PDF + JSON. The PDF is one-page summary plus per-dimension detail with the underlying tool citations.
mAb/ADC/oligo at launch. Bispecifics, peptide-drug conjugates, and protein degraders are on the roadmap.
Want to see all 4 modules in one place?Explore the full inscinstech.ai platform

Score a molecule. Save a quarter.

5 free mAb developability scores per month. No credit card.

Score nowBook a demo with a real molecule
InAnvil · Molecule design + manufacturability agent | inscinstech.ai