$SLNG is the name the AI power trade hasn’t priced yet. Yesterday, OpenAI announced Project Camellia — a $20 billion, 3.2-gigawatt data center campus in Effingham County, Georgia, with power from $SO‘s Georgia Power subsidiary phased in between 2028 and 2032. That timeline is the tell. Grid power ordered today doesn’t arrive until 2028 at the earliest, which means the compute capacity the AI economy needs right now has to be bridged some other way. Stabilis Solutions (Nasdaq: $SLNG), a Houston micro-cap with an ~$95M market cap, is already positioned in that gap.
In February 2026, Stabilis executed a multi-year, take-or-pay contract worth an estimated $200 million to supply LNG for behind-the-meter power generation at a U.S. data center. Deliveries run from Q1 2027 through Q1 2029 — exactly the window before gigawatt-class grid capacity comes online. The structure is iron-clad: $25 million in advance payments were agreed to fund logistics, of which $15 million was already received in Q1 2026, generating $12.4 million in operating cash flow even as headline revenue sagged. Management said on the Q1 call they were finalizing additional behind-the-meter data center opportunities expected to commence in Q2 2026 — a catalyst that has not yet shown up in any reported numbers.
The broader setup matters here. $GEV booked $2.4 billion in data center electrification orders in Q1 2026 alone, and $CAT is running power-generation revenue up 41% year-over-year. Those are multi-billion-dollar platforms priced for perfection. $SLNG is a different animal: 130,000 gallons-per-day of current LNG liquefaction capacity in Texas and Louisiana, a proven last-mile delivery network, and a contract backlog that represents more than twice its entire market cap in forward revenue. The Galveston LNG liquefaction project — 350,000 gpd of new capacity, total capex $350–400M — is the longer-dated call option. Carnival signed a 10-year marine bunkering agreement as anchor offtake; 56% of Galveston capacity is already committed.
Catalysts to watch: Q2 2026 earnings (expected August 2026) should confirm whether additional BTM data center contracts signed as guided; any Galveston FID announcement would be a step-change event; and the August 10, 2026 Annual General Meeting may bring strategic updates on the pipeline. Watch for incremental $15M advance payments from the primary data center customer as delivery prep ramps toward Q1 2027.
Risk: This is a micro-cap in transition. Q1 2026 revenue fell 40% year-over-year to $10.4M as two legacy contracts rolled off, and cost of revenues hit 96% of sales — margins are deeply compressed until new contracts ramp. The primary data center contract doesn’t generate meaningful revenue until 2027. Customer concentration is severe: Q1 2026 10-Q disclosed a single unnamed customer accounted for essentially all R&D/contract service revenue. The Galveston FID has already slipped past its original Q1 2026 target, and the $350–400M project-level financing remains unconfirmed. An equity shelf of up to $100M filed in March 2026 signals potential dilution ahead. This is a thesis stock, not a revenue story — yet.
Not financial advice.