MP1: Model Update Post Latitude.sh A$254M Contract

Research evidence excerpt

MP1: Model Update Post Latitude.sh A$254M Contract

are US-based, EBITDA: Company reported EBITDA (excl. restructuring charges) for FY23/FY24 was

institutionally backed AI application and inference providers, not A$25.2m/$57.1m.All values in AUD unless otherwise noted.

neoclouds. Agentic coding (GitHub runner-type workloads) is a demand Priced as of prior trading day's market close, EST (unless otherwise noted).

driver alongside inference.

• Funding/capex: US$101.0M capex funded via cash and a newly upsized

A$150.0M SOFR-plus bank debt facility. These contracts, combined with

the April contract and business-as-usual investment, satisfy the US

$86.0M CY26/CY27 capex requirement agreed at acquisition. GPU fleet:

GPUs are <10% of the existing fleet, with on-demand GPU capacity

effectively fully utilized. Contract-backed deployment is the preferred

approach for GPUs given capital requirements. Lower-power-draw

models (H100s, RTX) are deployed into traditional DC environments.

• Guidance/risks: FY26 revenue and EBITDA guidance reaffirmed. FY26

Group capex A$90-100M unchanged; could increase by up to A$140.3M

if hardware delivers before 30 June 2026. With contracted ARR of ~US

$73.6M secured, both the CY26/CY27 revenue milestones of US$74.2M/

US$115M appear within reach, reducing uncertainty around the US

$150M in contingent payments. Key risks include contract renewal (24-36

month terms), integration, hardware supply timing, and capex discipline.

Non-renewal could result in a step-down in contracted revenue with

assets reverting to the on-demand pool at uncertain utilization rates.

• RBC estimates: FY26e unchanged. FY27e compute revenue raised to A

$174M (from A$97M). FY27e group revenue increases to A$494M (from

A$410M). FY27e EBITDA increases to A$168M (from A$122M). FY27e