AI Summary
About
Exscientia was an AI-driven drug discovery and design company founded in Oxford, UK (spun out of the University of Dundee in 2012). It pioneered using machine learning to design small-molecule drug candidates, and in 2020 one of its candidates was reported as the first AI-designed molecule to enter human clinical trials. The company went public on Nasdaq in October 2021 under the ticker EXAI, raising about $510.4 million with backing from SoftBank and the Bill & Melinda Gates Foundation.
After a difficult 2023–2024, Exscientia merged with Recursion Pharmaceuticals in an all-stock deal valued at about $688 million that closed on November 20, 2024. Exscientia shareholders received 0.7729 Recursion shares each and ended up owning roughly 26% of the combined company; Recursion holders kept about 74%. The merged entity operates as Recursion (Nasdaq: RXRX) and runs Exscientia’s design technology inside the Recursion Operating System — a platform that pairs an automated wet lab generating millions of cell experiments a week with models trained on more than 50 petabytes of proprietary biological data.
The key thing to understand for pricing: this is not a SaaS company. It does not sell seats, API calls, or subscriptions. Its commercial model is the pharmaceutical industry’s classic R&D collaboration structure, applied to an AI platform.
For current corporate information, see Recursion and the legacy Exscientia site.
Pricing summary : How Exscientia’s pricing model works
There is no rate card, and the money largely flows toward the company, not away from it. Exscientia / Recursion monetizes its AI drug-discovery platform two ways:
-
Partnered programs. A large pharma company pays an upfront cash payment, funds the research, then pays milestone payments as a molecule clears each discovery, clinical, and regulatory gate, and finally pays tiered royalties on net sales once a drug is commercialized. Headline examples: Sanofi paid $100 million upfront with up to $5.2 billion in milestones (2022); Bristol Myers Squibb committed potentially more than $1.3 billion including up to $50 million upfront and up to $125 million in near-to-mid-term milestones; Merck KGaA paid $20 million upfront with up to $674 million in aggregate milestones (2023).
-
Internal pipeline. Recursion also develops wholly-owned drug candidates, bearing the cost and risk itself and capturing the full economic upside if a drug succeeds (or out-licensing it later). This is monetized as clinical-asset value, not as a recurring fee.
The combined partnered portfolio carries the potential for over $20 billion in milestone payments before royalties, with royalties ranging from mid-single-digit to double-digit percentages.
What makes this different: the “price” is outcome-based and back-loaded by design. A partner pays small amounts early and the large milestone and royalty payments only trigger if the AI-discovered molecule actually works in the clinic and reaches market. The platform’s value is realized over a 10-plus-year drug-development arc, not a monthly billing cycle.
Pricing by product
There are no tiers in the SaaS sense. The table below maps the three commercial “products” to how each is paid for.
| Offering | What you pay | Included | Key mechanics |
|---|---|---|---|
| Self-serve / seats | Not offered | — | No free tier, no trial, no public sign-up or API price |
| Pharma collaboration | Upfront fee + R&D funding + milestones + royalties | Joint program access to the AI design platform | Sales-led, partner-funded; payments back-loaded against clinical and regulatory milestones |
| Internal pipeline | Recursion’s own capital | Wholly-owned drug candidates | Value captured as clinical-asset / equity upside, not a fee |
Reference deal economics (reported in press releases and filings, not list prices):
| Partner | Upfront | Milestone potential | Royalties |
|---|---|---|---|
| Sanofi (2022) | $100 million | Up to $5.2 billion | High-single-digits to mid-teens (co-invest option to 21%) |
| Bristol Myers Squibb | Up to $50 million | Up to $125 million near-/mid-term; potentially more than $1.3 billion total | Tiered on net sales |
| Merck KGaA (2023) | $20 million | Up to $674 million aggregate | Mid-single-digits to low-double-digits |
| Bayer (Recursion) | Not disclosed | Up to $1.5 billion | On net sales |
Sales motions across products: fully sales-led and partner-led. There is no self-serve motion — every relationship is a negotiated business-development deal between two companies.
Hidden costs : What Exscientia partners actually pay
For a SaaS buyer the “hidden costs” question is meaningless here — there is no seat price to be surprised by. The real cost structure for a pharma partner is different in kind:
- R&D funding obligation. Beyond the upfront, the partner typically funds the research program itself — FTE costs, lab work, and trials — which dwarfs the upfront payment over the life of a program.
- Milestone exposure. The eye-catching “up to $5.2 billion” figures are contingent and cumulative across many milestones; most programs never trigger the full amount because most drug candidates fail. The headline number is a ceiling, not an invoice.
- Royalty stacking. Tiered royalties (mid-single-digits up to ~mid-teens, and up to 21% if Exscientia co-invested with Sanofi) are paid on net sales for the commercial life of any approved drug — a long-tail cost that only begins if a drug reaches market.
| Line item | Cost |
|---|---|
| Upfront cash payment | $20 million to $100 million (deal-dependent) |
| Ongoing R&D / FTE funding | Partner-funded, undisclosed, multi-year |
| Milestone payments | Triggered per stage; up to $5.2 billion ceiling on the largest deal |
| Royalties on net sales | Mid-single-digits to ~mid-teens (up to 21% with co-investment) |
Want to model an outcome-based or milestone deal structure? Use the Exscientia pricing calculator to sketch upfront-plus-milestone economics, and see the usage-based pricing guide and outcome-based pricing trends.
Pricing evolution : Exscientia pricing history and changes
Cadence
| Year | Milestone | Notes |
|---|---|---|
| 2021 | Nasdaq IPO (EXAI) | Raised ~$510.4M; SoftBank + Gates Foundation backing |
| 2022 | Sanofi mega-deal | $100M upfront, up to $5.2B milestones — largest deal to date |
| 2023 | Merck KGaA deal | $20M upfront, up to $674M aggregate |
| 2024 | Merger into Recursion | ~$688M all-stock; closed Nov 20, 2024 |
| 2025 | Recursion FY revenue ~$74.7M | Almost entirely milestone/collaboration income |
Tracked range: 2021–present, drawn from company press releases and SEC filings.
Notable changes
- Jan 7, 2022 — Sanofi strategic collaboration: $100 million upfront, up to $5.2 billion in milestones, royalties of high-single-digits to mid-teens, with an option to co-invest and lift royalties to 21%.
- Sep 19, 2023 — Merck KGaA collaboration: $20 million upfront, up to $674 million in aggregate milestones, royalties from mid-single-digits to low-double-digits.
- Nov 20, 2024 — Merger into Recursion completed (~$688 million all-stock). Exscientia ceases to trade independently; the platform becomes part of the Recursion Operating System.
- Oct 2025 — Recursion accepted a $30 million milestone from Roche & Genentech for a second neuroscience Phenomap, illustrating how milestone income lands lumpy and unpredictable quarter to quarter.
What’s unique : Exscientia’s distinctive pricing mechanics
1. The AI platform is monetized as drug-program economics, not software. Unlike almost every other company in this blueprint, there is no per-seat, per-token, or per-API meter. The “unit” of value is a drug program and the payment schedule is the pharma industry’s upfront-plus-milestones-plus-royalties structure layered on top of an AI engine.
2. Pricing is radically outcome-based and back-loaded. A partner pays tens of millions upfront against billions in contingent milestones that only trigger if the molecule survives the clinic. The vendor effectively shares the 90% failure rate of drug discovery — the headline “biobucks” numbers are ceilings tied to success.
3. The company can be the customer. Through its internal pipeline, Recursion funds its own programs and captures the full upside rather than charging a third party. That dual model — partner-funded and self-funded — is unusual and means “revenue” and “price” partly decouple from any external customer.
Strengths & weaknesses
| Strengths | Weaknesses |
|---|---|
| Outcome alignment — partners pay big only when a drug actually advances | Revenue is lumpy and unpredictable; a single $30M milestone can swing a quarter |
| Enormous upside ceilings (Sanofi up to $5.2B; portfolio >$20B in milestones) | Most milestone potential never converts — drug attrition is ~90% |
| Validated by tier-one partners (Sanofi, BMS, Merck KGaA, Roche/Bayer) | No self-serve revenue; growth gated by slow, sales-led BD cycles |
| Proprietary data moat (>50PB) + BioHive-2 compute hard to replicate | Opaque to outsiders — no list price, long path to commercial royalties |
Billing UX : Exscientia billing controls and transparency
- Billing controls — None in the SaaS sense. “Billing” is contractual: a master collaboration agreement defines upfront, milestone, and royalty schedules negotiated by business-development and legal teams.
- Usage visibility — There is no usage dashboard or spend meter. A partner tracks program progress against milestone gates, not API consumption.
- Payment options — Payments are structured cash transfers and royalty accruals governed by the deal, recognized as collaboration revenue in the partnered company’s financials (Recursion reported ~$74.7 million of such revenue in FY2025).
Strategic wins : Why Exscientia’s pricing decisions worked
1. Selling outcomes, not access
By pricing as milestones-plus-royalties rather than a platform license, Exscientia aligned its economics with the partner’s only real goal — an approved drug — and unlocked deal ceilings (up to $5.2 billion with Sanofi) that no software seat price could ever reach. This is the purest form of outcome-based pricing: the vendor gets paid as value is proven.
2. Anchoring credibility through tier-one partners
Landing Sanofi, Bristol Myers Squibb, Merck KGaA, and (via Recursion) Bayer and Roche/Genentech turned the AI platform’s claims into validated demand. Each marquee deal de-risked the next and supplied non-dilutive capital. Related reading: how AI companies structure pricing.
3. Keeping internal upside
Running its own pipeline alongside partnerships meant Exscientia / Recursion did not have to sell every program to monetize the platform — it could keep the most promising assets and capture full drug value. See choosing the right usage metric for how picking the value unit (here, a drug program) shapes the whole model.
Areas to improve : Gaps in Exscientia’s pricing approach
1. Revenue is lumpy and hard to forecast
Because income arrives as discrete milestones, results swing wildly — Recursion’s Q4 2025 revenue jumped to ~$35.5 million from ~$4.5 million a year earlier purely on the timing of a single $30 million Roche/Genentech milestone. That volatility is the flip side of cost unpredictability, felt by the vendor instead of the buyer.
2. The headline “biobucks” overstate reality
Quoting “up to $5.2 billion” sets expectations that the ~90% drug-attrition rate rarely meets. More transparent probability-weighted framing would build trust, but pharma norms keep the ceiling numbers front and center.
3. No on-ramp for smaller collaborators
The fully sales-led, mega-deal-only model leaves no path for smaller biotechs or academic groups to access the platform at a lower commitment. A tiered or limited self-serve research offering could widen the funnel without diluting the flagship partnerships.
Key takeaways
- There is no rate card. Exscientia / Recursion prices AI drug discovery as a pharma R&D collaboration — upfront fee, milestone payments, and royalties — not as software.
- Outcome alignment unlocks huge ceilings. Tying price to drug-development milestones produced deal values (up to $5.2 billion with Sanofi) impossible under any seat or usage meter.
- The numbers are contingent. Headline “biobucks” are ceilings gated by clinical success; most never convert because most drugs fail.
- Exscientia no longer exists independently. It merged into Recursion (Nasdaq RXRX) in an ~$688 million all-stock deal that closed November 20, 2024.
- Lumpy revenue is the cost of outcome pricing. Milestone-driven income is unpredictable quarter to quarter — a structural trade-off of selling outcomes rather than access.
UBP implications
- Outcome-based pricing is the logical endpoint when the “unit” is a proven result. Exscientia shows what pure outcome pricing looks like — payment tied entirely to milestones the customer cares about, with the vendor absorbing failure risk.
- Back-loaded pricing trades predictability for ceiling. Teams adopting milestone or success-fee models should expect the same volatility Recursion lives with, and plan cash flow accordingly.
- Pick the value unit that matches what the buyer actually buys. For drug discovery that is a program, not an API call — a reminder that the right meter follows the customer’s outcome, not the vendor’s infrastructure. See choosing the right usage metric.
Sources
- Recursion — Partner with Recursion (accessed 2026-06-16)
- Recursion / Exscientia merger completion announcement (accessed 2026-06-16)
- Exscientia & Sanofi strategic collaboration (accessed 2026-06-16)
- Exscientia & Merck KGaA collaboration (accessed 2026-06-16)
- Recursion FY2025 financial results (accessed 2026-06-16)
- Cooley — Exscientia $510.4M IPO (accessed 2026-06-16)
Bottom line
Exscientia is the clearest case in this blueprint of pricing an AI platform as outcome economics rather than software. It charges nothing per seat or per call; instead, pharma partners pay an upfront fee, fund the research, and owe milestone payments and royalties only as an AI-discovered drug proves itself — Sanofi alone committed up to $5.2 billion against a $100 million upfront. That model is now part of Recursion (Nasdaq RXRX), which absorbed Exscientia in an ~$688 million all-stock merger in November 2024. The lesson for usage-based pricing: when the unit of value is a proven result, the meter follows the outcome — and the vendor shares the risk.
Want to compare Exscientia against other outcome-priced and enterprise AI companies? Browse the pricing blueprint.
Pricing timeline : Major events on a vertical axis
Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.
Merger into Recursion completed
All-stock deal valued at about $688M closed; Exscientia folded into Recursion Pharmaceuticals (Nasdaq RXRX). Exscientia holders received 0.7729 Recursion shares each and ~26% of the combined company. The merged partnered portfolio carries over $20B in potential milestones before royalties.
Merck KGaA collaboration
$20M upfront plus up to $674M in aggregate discovery, development, regulatory, and sales milestones across three initial programs, with tiered royalties from mid-single-digits to low-double-digits.
Sanofi strategic collaboration
Exscientia's largest deal: $100M upfront, up to $5.2B in milestones, tiered royalties of high-single-digits to mid-teens, with an option to co-invest and raise its royalty to 21%. Covered up to 15 small-molecule candidates across oncology and immunology.
- · Exscientia's drug candidate for OCD (DSP-1181, with Sumitomo Dainippon) was widely reported as the first AI-designed molecule to enter human clinical trials, in 2020 — designed in under 12 months versus a typical multi-year cycle.
- · Exscientia raised about $510.4 million in its October 2021 Nasdaq IPO (ticker EXAI), with concurrent private placements from SoftBank and the Bill & Melinda Gates Foundation — one of the largest biotech IPOs of that year.
- · Its 2022 Sanofi deal carried up to $5.2 billion in potential milestones — larger than the entire $688 million all-stock value of the company's own 2024 merger into Recursion.
Questions & answers
- What is Exscientia's pricing model?
- Exscientia (now part of Recursion) does not sell software seats or API credits. It is an AI-driven drug discovery company, so its 'pricing' is the economics of pharma R&D collaborations: a partner pays an upfront fee, funds the research, then pays milestone payments as a molecule advances through discovery, clinical, and regulatory stages, plus tiered royalties on eventual product sales. Every deal is negotiated individually — there is no list price.
- Does Exscientia offer a free tier?
- No. There is no free tier and no self-serve plan. Access to the platform comes only through a negotiated, sales-led partnership or co-development agreement. The company also runs its own internal drug pipeline rather than licensing the platform to small customers.
- How much does an Exscientia partnership cost?
- Partnerships are quoted, not listed, and the money flows toward the partner-funded company rather than away from it. As reference points, Sanofi paid $100 million upfront with up to $5.2 billion in milestones (2022), Bristol Myers Squibb committed potentially more than $1.3 billion, and Merck KGaA paid $20 million upfront with up to $674 million in aggregate milestones (2023). All include tiered royalties on net sales.
- Is Exscientia still an independent company?
- No. Exscientia merged with Recursion Pharmaceuticals in an all-stock deal valued at about $688 million that closed on November 20, 2024. The combined company operates as Recursion (Nasdaq: RXRX); Exscientia's former Nasdaq ticker EXAI no longer trades. The technology now runs inside the Recursion Operating System platform.