RXRX Stock: Why Did Nvidia Offload AI Drug Developer Recursion?

Based in Salt Lake City, Utah, Recursion Pharmaceuticals [RXRX] is a clinical-stage biotechnology firm. It was founded in 2013 to industrialise drug discovery using automation, high-throughput biology and machine learning, positioning itself early within the “TechBio” category. Emerging from academic roots at the University of Utah, it attracted venture backing and strategic pharma partnerships, including Sanofi’s [SNY] rare disease unit Genzyme and later Bayer [BAYRY].

The firm listed on the Nasdaq in April 2021; this was followed by a $12bn collaboration with Roche [RHHBY].

Since then, the trajectory has been volatile but narrative-driven: Nvidia’s [NVDA] $50m investment and artificial intelligence (AI) positioning drove sharp rallies, while clinical uncertainty and persistent losses have weighed on sentiment.

The RXRX share price has crumbled in the years since its IPO – as of March 25, it is 82% down from its first opening price, and more than 22% down in the year to date. It might not all be doom and gloom, however. 

Let’s go into more depth on some recent developments, and tease out an investment case for RXRX stock.

Nvidia Partnership: Highs and Lows 

Back in 2023, Nvidia committed $50m to a multi-year collaboration with Recursion to develop AI-driven software aimed at improving patient treatments.

Recursion also outlined plans to accelerate the development of its AI foundation models across biology and chemistry, with the intention of optimising and distributing them to biotech partners via Nvidia’s cloud infrastructure.

The company said it would leverage its proprietary dataset – spanning more than 23 petabytes and trillions of gene-compound relationships – to speed up model training on Nvidia DGX Cloud.

Recursion already operates its in-house supercomputer, BioHive-1, which is powered by Nvidia’s advanced GPUs and underpins its broader push to industrialise AI-led drug discovery.

“The collaboration between Recursion and Nvidia marks a significant milestone on our journey toward revolutionising drug discovery through AI,” said Recursion’s Chief Technology Officer Ben Mabey in a statement.

Recursion hoped its use of AI would give it a competitive edge, allowing it to develop and commercialise medicines faster than its peers. However, this did not come to pass. It has still not launched a single drug, and its perceived advantage has eroded as rivals increasingly adopt similar approaches.

This may be why, on 19 February 2026, Nvidia announced in its latest 13F filing that it had completely exited its position in RXRX, despite the fact that their partnership was still ongoing. Hardly a vote of confidence.

Nonetheless, other major players in the tech investment space did not share Nvidia’s stance on the stock. Soon after the disclosure, Cathie Wood moved to increase exposure. ARK Invest [ARKK] acquired 1.25 million shares of Recursion across two ETFs, building on a position it had been accumulating in recent months.

What did Wood see that Nvidia didn’t? The answer might be found in Recursion’s most recent earnings. 

Earnings Breakdown

Reporting at the end of February, Recursion significantly beat Wall Street expectations in its Q4 2025 results.

The company reported $35.5m in Q4 revenue, mostly from collaboration income, up from $4.5m a year earlier and roughly $11m above consensus. The increase was largely due in part to a $30m milestone payment received from Roche in October.

For the full year, total revenue rose approximately 27% year-on-year to $74.7m, boosted by the inclusion of financials from its 2024 Exscientia acquisition.

Net loss per share reached $0.21, $0.08 better than expected, due to R&D and SG&A costs falling 2% and 56% y/y to $95.9m and $33.7m, respectively.

Liquidity stood at $743.3m in cash and equivalents at year-end, up from $594.4m a year ago, giving the company a projected runway into 2028 without extra financing.

Pipeline Going Forward

As Simply Wall St. recently noted, to buy into Recursion is to “buy into the idea that its AI-native Recursion OS, supercharged by the Nvidia-backed BioHive-2, can turn massive experimental datasets into differentiated drugs before cash runs out.”

The company knows this. 

On the Q4 earnings call, CEO Najat Khan underlined that “Recursion has reached an inflection point: moving from proving that AI can participate in drug discovery to demonstrating that an AI-native operating system can generate clinical proof and durable value. 

“We enter 2026 with five differentiated clinical programmes advancing with defined next-stage milestones, a growing discovery portfolio informed by proprietary platform insights and more than $500m in upfront and progress-based milestone payments earned to date.”

This is seemingly a vision that Cathie Wood shares. 

Competitor Comparison: ABCL vs SDGR vs RXRX

To get a sense of how near or how far Recursion is from its goals, let’s line it up alongside two other firms in the space.

AbCellera Biologics [ABCL] is a clinical‑stage biotech focused on AI‑enabled antibody discovery and development. It uses deep learning and computational platforms to screen and identify high‑potential therapeutic antibodies across multiple indications. The company has strategic partnerships with major pharma firms and is advancing several programmes through Phase 1/2 clinical trials, including menopausal symptom and autoimmune treatments, supported by steady pipeline progress. Unlike Recursion’s broad AI‑driven small molecule and biology platform, AbCellera specialises in biologics and antibody modalities, giving it a distinct focus within the broader AI‑biotech sector. ABCL remains development‑stage with increasing collaborations and early clinical catalysts.

Schrödinger [SDGR], meanwhile, operates at the intersection of computational chemistry, physics‑based modelling and AI to accelerate molecular discovery. Its platform supports both internal preclinical programmes and software licensing to pharma/biotech partners, diversifying revenue beyond pure drug development. Though the stock has lagged over the past year and remains unprofitable, it reported growing revenue and maintains a broad suite of computational tools used across therapeutic areas. Schrödinger’s model contrasts with Recursion’s experimental wet‑lab AI platform by emphasising in‑silico design and simulation, reducing reliance on traditional experimental throughput.

Source: Yahoo Finance

From an investor’s standpoint, Recursion, AbCellera and Schrödinger represent distinct risk-reward profiles within AI‑linked biotech. 

Recursion’s end‑to‑end platform, expanded through the Exscientia integration, offers the most comprehensive AI drug discovery pipeline but remains execution‑dependent on clinical readouts and internal programme success. 

AbCellera’s focus on antibody discovery and strong partnerships gives it a clearer near‑term path to milestones with potentially less binary risk than single‑asset clinical bets, albeit still dependent on R&D outcomes. 

Schrödinger’s hybrid model, combining software licensing with internal programmes, provides diversified revenue and may offer a more stable cash flow foundation, though growth is tied to broader adoption of computational platforms. Investors must weigh platform breadth (Recursion) against modality focus (AbCellera) and software‑plus drug discovery balance (Schrödinger).

Conclusion: The Investment Case for RXRX Stock

Recursion offers a high-risk, high-reward opportunity in AI-driven drug discovery. 

Bullish arguments point to its vast proprietary dataset, Exscientia acquisition, BioHive supercomputer, and partnerships with Roche and Nvidia, providing a platform capable of industrialising drug discovery. A Q4 2025 revenue beat, a projected cash runway to 2028 and Cathie Wood’s buy-in all support confidence in long-term upside. 

Bear cases highlight the fact that Recursion has yet to launch a single drug, that its competitive edge is eroding as peers adopt AI, that Nvidia exited its position and that past stock performance has been extremely weak.