Optibrium is one of the very few vendors in computational drug discovery that will put a number in writing, and the number is this: StarDrop annual subscriptions typically start at around $10,000 and can rise into the six figure range depending on the breadth of function and the number of users. That figure comes from Optibrium's own knowledge base, not from a directory listing or a summary of one, which already makes it more reliable than almost anything else published about this category.
What the sentence leaves open is the spread. A factor of ten between the floor and the ceiling is not vendor evasiveness, it is the licensing model doing its job: StarDrop is a core product with ten optional modules hanging off it, sold per user, per site or as a floating pool. This piece takes that apart so you can predict roughly where in the range your quote will land before you spend three weeks finding out.
What the base subscription includes
The core of StarDrop is a data visualisation and design environment rather than a calculation engine. Out of the box it does interactive charts and chemical spaces, structure activity analysis across a series, multi parameter optimisation to balance potency against ADME and physicochemical properties, and two of the vendor's signature views: Card View for relationships inside a chemical series, and Glowing Molecule for visual feedback on where a design change is helping or hurting.
Optibrium's pricing page is explicit that several things people expect as line items are included rather than extra: deployment on the desktop or in the cloud, tailored hands on training delivered on site or virtually, ongoing email and phone support with responses typically within a few hours, the video and worked example library, and updates to new features as they ship. That matters when you compare quotes. A cheaper headline number from a vendor that prices training and support separately is not cheaper.
What moves the price from five figures to six
Optibrium lists four drivers, and they show up in almost every quote in this market. The number of users comes first, split across single user, site specific, company wide and floating concurrent licences. Breadth of capability comes second, meaning how many of the ten optional modules you switch on. Hosting is third: cloud hosting by the vendor carries an additional cost but can lower total cost of ownership against running it yourself. Customisation and integrations are fourth, covering collaborative functionality, database access and connections to third party software.
The floating licence is the lever most small teams underuse. It buys concurrent sessions rather than named people, so a group of six chemists who each touch the software a few hours a week can share a much smaller pool. If a quote arrives priced per named user and your usage is bursty, asking for the floating shape by name is the single cheapest change you can make to the number.
The ten modules, and which ones a quote usually turns on
Here is the module set as Optibrium lists it. The category column is worth reading, because the three groupings buy very different things and a team rarely needs more than one of them at the start.
| Module | Category | What it adds |
|---|---|---|
| ADME QSAR | In silico modelling | Predictive models for key ADME and physicochemical properties |
| Metabolism | In silico modelling | Identifies metabolic pathways and guides design away from liabilities |
| Derek Nexus | In silico modelling | Knowledge based prediction across 40 or more toxicity endpoints |
| Auto-Modeller | In silico modelling | Builds and validates QSAR models on your own chemistry |
| Nova | Generative chemistry | Generates and prioritises new compounds from virtual libraries |
| BIOSTER | Generative chemistry | Extends Nova with 29,000 or more structure modifications |
| Inspyra | Generative chemistry | Combines expert chemistry knowledge with generative methods |
| eSim 3D | 3D design | Ligand based 3D design, binding conformations and actives |
| SeeSAR | 3D design | Structure based design, visualising ligand protein interactions |
| MPO Explorer | Strategy | Builds multi parameter optimisation strategies by rule induction |
For an antibacterial program the honest reading of that table is mixed. ADME QSAR, Derek Nexus and Metabolism are mature and genuinely useful once you have a series worth developing. Auto-Modeller is the module people reach for when they want their own potency model, and it carries the same requirement every model builder does, which is a consistent in-house dataset to fit on. The generative and 3D design modules assume structure based work against a defined target, which is the shape of a lot of antibacterial research but not of whole cell screening.
How much does StarDrop cost for a small team?
Realistically, the bottom of the published range. About $10,000 a year is the entry point for the core product with a small number of seats and little or no module content, and each module and each additional named user pushes it upward from there. Optibrium says the six figure end applies to global site licences with several modules, which is a large pharma shape rather than a four person lab.
Worth knowing for context: Optibrium publishes separate figures for its other products, and they set the ceiling of this market. Its BioPharmics 3D modelling subscription runs $100,000 to $250,000 a year by organisation size, with an all inclusive model covering unlimited users and CPU rather than per token billing. Cerella, the AI discovery platform, is described as getting into the $100,000 and up ballpark. The category range Optibrium publishes for 3D molecular modelling software generally is $50,000 to $1,000,000 a year. Against those, StarDrop is the accessible product in its own catalogue.
Is there a free version of StarDrop?
No. Optibrium offers a free demonstration and a conversation with its team, which is not the same thing as a free tier you can work in. That is the norm for this category rather than a mark against it: the comparable vendors either offer nothing, or offer a viewer that cannot run a calculation. The practical consequence is that evaluating StarDrop means booking time with the vendor, which is fine for a planned procurement and slow when you need an answer this week.
What to ask before the quote lands
Four questions save the most money. Ask for the quote itemised by module and by seat, so you can see what you would actually be paying for the parts you use. Ask explicitly about floating concurrent licences against named users. Ask what the second year costs, because renewal is where multi year discounting shows its teeth. And ask whether the term can be aligned to a grant cycle, since a one year term priced slightly higher is the right trade when the science may pivot.
If the deployment has to satisfy an auditor, add a fifth. Validation documentation, qualified environments and change control are priced separately by most vendors in this space, and they are the item most often discovered after the budget was fixed. A regulated program is better off tracking those obligations in a proper compliance program than assuming the software contract will carry them.
When StarDrop is the wrong instrument
A design environment earns its subscription when there is a series to design and someone whose job is to design it. If the recurring question in your program is narrower, which compounds from this library are worth plate time against which organisms, then most of what you would be buying sits unused. That is not a criticism of the product. It is a mismatch between a broad instrument and a single repeated question.
For that narrower question the arithmetic is different in kind. Screening here is $1,788 a year at the entry plan and $17,880 a year at the largest one, published on the page, billed monthly, with no quote cycle in front of it. You paste a structure or a peptide sequence, pick a strain panel, and get a predicted MIC band per organism with the expected resistance mechanism named and a confidence level on every row. The full category comparison, including which vendors disclose anything at all, is in drug discovery software cost, and the equivalent breakdown for the other big suite is in Schrodinger software cost.
The Auto-Modeller question deserves its own note, because it is the specific reason antibacterial teams look at StarDrop. Building your own potency model is a real option and a real project: consistent assay data, one model per organism, an applicability domain to defend and someone to maintain it. The trade between building that and using a model that already exists is worked through in QSAR software for antibacterial activity, and the library scale version of the same decision sits in compound library screening software.
The short version
StarDrop starts at roughly $10,000 a year and climbs with users, modules, hosting and integrations, with the six figure end reserved for company wide deployments carrying several modules. Training, support and deployment are inside the price rather than beside it. There is no free tier, so evaluation runs through the vendor. And if your program's repeated question is triage against a panel rather than design across a series, a narrow subscription answers it for a fraction of the entry point of any design suite, including this one.
Vendor pricing moves. Everything above was read at Optibrium's own pages at the time of writing, and the figures should be confirmed with the vendor before a budget depends on them. Anything you read about this market in a directory summary, including numbers attributed to vendors who publish none, deserves more suspicion than it usually gets.