Nebius Faces a High-Stakes Earnings Test With a $643 Million AI Bet Already in Play

Nebius reports Q1 results May 13 with $50B backlog, two AI acquisitions in three months, and options pricing a near-8% move. Analysts see 578% revenue growth but widening losses.

The options market is bracing for a near-8% swing when Nebius Group reports first-quarter results on May 13 before the opening bell — and history suggests the actual move could be far larger. In August 2025, the stock surged more than 27% despite derivatives pricing in only a single-digit jolt. With shares now trading just shy of their all-time high near $196, the bar for a positive surprise sits uncomfortably high.

A Second Acquisition in Three Months Reshapes the Narrative

Just weeks before the earnings release, Nebius dropped $643 million on Eigen AI, a startup founded by MIT HAN Lab researchers with only 20 employees — a price tag of roughly $32 million per head. The deal, structured as $98 million in cash plus 3.8 million Class A shares with vesting conditions for key personnel, marks the company's second acquisition in three months. In February, Nebius bought Tavily, an AI agent search provider, for $275 million.

Both transactions push the Dutch neocloud provider — which spun off from Yandex in 2024 — away from its roots as a pure GPU lessor and toward higher-margin platform services. Eigen AI specializes in post-training optimization for open-source models, including techniques like post-training quantization, KV-cache optimization, and custom CUDA kernels. The startup's optimization layer will plug directly into Nebius' Token Factory, which offers scalable inference endpoints and fine-tuning pipelines. The combined technology has already produced optimized versions of DeepSeek, Llama, and Qwen that hit output speeds of up to 911 tokens per second in benchmarks.

The Eigen AI team is also tasked with establishing a Nebius presence in the San Francisco Bay Area, giving the company a foothold in the heart of the AI ecosystem.

The Backlog Is Massive — But So Is the Cash Burn

Nebius enters earnings with a contracted order backlog approaching $50 billion, anchored by a multiyear deal with Meta worth $27 billion and a Microsoft commitment of up to $19.4 billion. NVIDIA has made a strategic $2 billion investment, and a $4.34 billion convertible note secures the financing for the company's expansion plans.

Yet the infrastructure buildout is devouring capital at a staggering pace. Nebius plans capital expenditures in the double-digit billions for 2026 alone, and analysts estimate the company will not generate positive free cash flow until 2030. A significant portion of expected revenue is back-end loaded to the second half of the year, amplifying execution risk.

What Wall Street Wants to See

For the first quarter, analysts project revenue of roughly $375 million to $389 million — representing growth of more than 578% to 600% year over year. The loss per share is expected to widen to $0.77 from $0.41 in the prior-year period. Eight of ten Wall Street analysts rate the stock a buy, with price targets ranging from Bank of America's revised $175 (up from $150 on new contract wins) to BWS Financial and DA Davidson both calling for $200.

The critical metric, however, is annualized recurring revenue. Management has guided for an average ARR of $8 billion for the full year 2026, with a target range of $7 billion to $9 billion. The first-quarter ARR trajectory will signal whether that ambition is realistic. Investors also want concrete updates on infrastructure projects with Meta and Microsoft, as well as early integration details for Eigen AI.

A Structural Shift in the AI Computing Market

The timing of the Eigen AI acquisition reflects a broader transformation in the AI computing landscape. Demand for inference compute — the processing power needed to run AI models in production — is poised to overtake demand for training compute. Reports of ongoing discussions between Nebius and AI developer Anthropic fit neatly into this thesis, as inference workloads require precisely the kind of optimization Eigen AI provides.

In a market with transparent pricing and low switching costs, the ability to offer either lower cost per token or higher throughput from the same hardware gives Nebius a competitive edge. That edge will be tested against hyperscalers and rival neocloud providers as the company scales.

The Two-Way Risk

With the stock trading near its all-time high and well above its 200-day moving average, the downside risk is substantial if the May 13 report disappoints. A weak outlook on Eigen AI integration or delays in hyperscaler projects could trigger a sharp correction. Conversely, a strong quarter with convincing ARR momentum could push shares decisively above the $200 mark.

The options market has priced in volatility, but the pattern of recent earnings surprises suggests the actual move could exceed expectations — in either direction.