Trans-Caspian Cable: The Construction Is Finished. The Launch Date Is Still Moving.

The first fibre-optic cable across the Caspian Sea is built. Whether it carries commercial traffic this month or in December depends on which official you asked, and when.

On 2 October, NEQSOL Holding announced that the construction phase of the Trans-Caspian Fiber Optic Cable Line was complete, that testing was under way, and that launch and the start of operations were planned for October 2026.

On 6 October, Rashad Hasanov, Azerbaijan's Deputy Minister of Digital Development and Transport, said the cable was scheduled to become operational by the end of 2026.

Four days apart, two statements about the same asset, roughly three months apart in substance. Neither is a retraction of the other, and neither side has announced a delay. But the distance between them is the most informative thing published about this project all year.

A target that has moved once already

The original commissioning date was the third quarter of 2026. Kazakhstan's Deputy Prime Minister and Minister of AI and Digital Development, Zhaslan Madiyev, set it out at a year-end briefing in December 2025, and repeated it in April and again at the Freedom Inside 2026 forum. Q3 ended on 30 September.

By August the language had already loosened: an Azerbaijani project partner described the system as scheduled to be fully operational by the end of 2026. Hasanov's 6 October comment is consistent with that revision. NEQSOL's October statement is the outlier on the optimistic side.

The likeliest reading is mundane rather than dramatic. Construction finishing and a system entering commercial service are different milestones, and the gap between them is filled by testing, integration with terrestrial infrastructure and customer provisioning. A company announcing completed construction and a ministry forecasting commercial operation can both be accurate and still name different months.

What was actually built

The line runs 380 kilometres across the Caspian seabed between Sumgait in Azerbaijan and Aktau in Kazakhstan. It is jointly owned by AzerTelecom, the wholesale carrier within NEQSOL Holding, and Kazakhtelecom, Kazakhstan's largest operator. Total investment is roughly $50.6 million.

The specification is more interesting than the headline capacity. HMN Tech supplied a 32-fibre-pair cable with remote optically pumped amplifiers, rated at up to 400 terabits per second, with US-based Pioneer Consulting supervising. The cable-laying vessel left Baku in late July; the cable reached the Kazakh shore in August.

The number that matters commercially is not the 400 Tbps. It is 3.7 milliseconds — the round-trip latency between the two coasts, and the lowest physically achievable across the basin.

Why latency, not bandwidth, is the business case

Bandwidth across Eurasia is not scarce. Routes are. Data between China, Central Asia and Europe currently travels north through Russian terrestrial networks or south through the congested Red Sea and Mediterranean corridor. Both have become harder to underwrite, politically and physically.

The Caspian crossing is a third path, and it is the data-layer twin of the Middle Corridor — the same bet on a route that touches neither Russian territory nor the Hormuz and Red Sea chokepoints, one layer up from freight.

Low, predictable latency is increasingly a precondition rather than a premium feature. AI inference workloads, hyperscaler peering and cross-border compute scheduling are all sensitive to it in ways that bulk transit traffic is not. A route that is marginally cheaper does not win that business; a route that is measurably faster can.

The second cable

The Caspian line is one half of a wider position. At the other end of the same corridor, Vodafone Ukraine — also part of NEQSOL Holding — is building Kardesa with Vodafone Group, a submarine cable system across the Black Sea worth roughly EUR 100 million. Its first landing is planned in Bulgaria in 2027, followed by Türkiye, Georgia and Ukraine.

Read together, the two projects describe a single thesis: own the physical layer that determines how Europe–Asia traffic is routed, at both water crossings, before the routing patterns settle. Georgia sits at the western terminus of any build-out that connects them.

What could go wrong

Three things, in order of near-term probability.

Commissioning slips again. The target has already moved from Q3 to year-end. A second slip would push first revenue into 2027 and start to matter to the return profile of a $50.6 million asset.

The environment is hostile to hardware. The Caspian's variable salinity, shallow coastal zones and seismic activity forced heavily armoured cable with enhanced tensile strength. The sea is also shrinking — levels have been falling by up to 30 centimetres a year — which complicates shore approaches over the asset's life.

Demand has to show up at the assumed price. The latency argument is sound in principle. It depends on hyperscalers and AI operators actually routing workloads through Baku and Aktau rather than treating the corridor as a backup path. A planned second leg to Turkmenistan remains at an earlier stage of planning.

What to watch

One thing: a commissioning announcement from AzerTelecom or Kazakhtelecom naming a live date and first customers. Until that exists, construction completion is an engineering milestone, not a revenue event. If it lands in October, NEQSOL's timeline was right. If it lands in December, the ministry's was.

Further Reading

• Azerbaijan and Kazakhstan's Digital Silk Road: 400 Tbps Caspian Internet Bridge Set for 2026 Launch
• NEQSOL Holding: group structure and coverage
• The Middle Corridor: the Trans-Caspian trade route

Sources: NEQSOL Holding statement, 2 October 2026; remarks by Deputy Minister Rashad Hasanov, 6 October 2026; briefings by Deputy Prime Minister Zhaslan Madiyev, December 2025 and April 2026; Interfax; The Astana Times; Submarine Networks.