Timeweb Cloud: one in three documentation requests comes from AI agents, not humans
Mikhail Shpakov, head of development at Timeweb Cloud, says the cloud has gained a new type of user: not humans, but AI agents. One in three requests to the provider's documentation already comes from AI vendors' networks, and servers are increasingly created not from the control panel but through the API, at an assistant's request.
AI-processed from Habr AI; edited by Hamidun News
On July 23, 2026, Mikhail Shpakov, head of development at Timeweb Cloud, published a column on Habr arguing that the cloud provider has gained a new type of user: AI agents. One in three requests to Timeweb Cloud's documentation now comes from AI vendors' networks, and servers are increasingly created through the API on an assistant's command rather than from the control panel.
Who uses the cloud now?
The new Timeweb Cloud user is not a human but an AI assistant acting at a human's request. According to Shpakov, agents first consult the documentation to understand how the API works, then build the requests that create servers themselves.
"Our cloud has a new type of user, and they are not people," —
Mikhail Shpakov, head of development at Timeweb Cloud.
Customers used to click buttons in the control panel. Now an assistant layer sits between the human and the infrastructure, translating a task like "spin up a server for me" into concrete API calls. The control panel does not disappear, but it stops being the sole and primary entry point: a model that reads the documentation and calls endpoints on the user's behalf takes over a significant share of the work.
What do Timeweb's numbers show?
One in three requests to Timeweb Cloud's documentation comes from AI vendors' networks. That is the column's key figure. In other words, documentation is increasingly read not by developers but by models, which either train on it or answer users' questions about how a given cloud works in real time.
- One in three requests to Timeweb Cloud's documentation comes from AI vendors' networks
- Servers are increasingly created via API requests rather than clicks in the control panel
- API requests are written by an assistant at a human's request
- The column's author is Mikhail Shpakov, head of development at Timeweb Cloud
- The team previously described an automated testing system with 5000+ checks and a visualization of the cloud
Shpakov stresses that he views this trend from inside the provider and backs his conclusions with internal figures rather than general reasoning about the market. By his account, there are many trends in the cloud market, but the agent-driven one is the one he sees most closely, and he considers it the most important for the coming years.
Why this matters to providers
For cloud providers, the shift toward agents changes product priorities: documentation and APIs become an interface not for humans but for machines. If one in three documentation requests comes from an AI network, then API usability and the accuracy of endpoint descriptions directly affect how well an agent can manage infrastructure.
Timeweb Cloud draws conclusions from this picture for the next few years: the programmatic layer that assistants access is becoming more important than the graphical interface. It is no longer enough for a provider to build a convenient panel; its API and documentation must be "understood" by someone else's model, which creates servers on the user's behalf.
What counts as a good product is changing, too. A polished dashboard stops being the deciding argument if real operations (creating servers, configuring networks, scaling) are initiated not by a person with eyes and a mouse but by an agent through a strictly documented API. The emphasis shifts toward machine-readable documentation and predictable endpoint behavior.
What this means
The cloud is turning into infrastructure that is increasingly managed not by people directly but by AI agents through APIs. For providers, this means the quality of documentation and API interfaces becomes a competitive advantage in a world where an assistant spins up the server and the human merely states the task in words. The figure "one in three documentation requests comes from AI vendors' networks" is an early but telling signal that this shift is already under way and is not just a forecast on paper.
- Oct 24, 2026GPT-6.1 Sol in Devin Scores 60.4% on FrontierCode at $0.31 per Task
- Oct 24, 2026Cognition launches Devin Oncall: 100% on SREGym and a production fix in 2m 21s
- Oct 24, 2026Goodfire launches "internal" monitors for AI agents that cost less than LLM oversight
- Oct 24, 2026Bending Spoons CEO: AI Writes 90% of Code, While Cheap SaaS Goes Under the Hammer
❓ Key Questions Answered
Who uses the cloud now?
The new Timeweb Cloud user is not a human but an AI assistant acting at a human's request. According to Shpakov, agents first consult the documentation to understand how the API works, then build the requests that create servers themselves. "Our cloud has a new type of user, and they are not people," — Mikhail Shpakov, head of development at Timeweb Cloud. Customers used to click buttons in the control panel. Now an assistant layer sits between the human and the infrastructure, translating a task like "spin up a server for me" into concrete API calls. The control panel does
What do Timeweb's numbers show?
One in three requests to Timeweb Cloud's documentation comes from AI vendors' networks. That is the column's key figure. In other words, documentation is increasingly read not by developers but by models, which either train on it or answer users' questions about how a given cloud works in real time. One in three requests to Timeweb Cloud's documentation comes from AI vendors' networks Servers are increasingly created via API requests rather than clicks in the control panel API requests are written by an assistant at a human's request The column's author is Mikhail Shpakov,
Why this matters to providers?
For cloud providers, the shift toward agents changes product priorities: documentation and APIs become an interface not for humans but for machines. If one in three documentation requests comes from an AI network, then API usability and the accuracy of endpoint descriptions directly affect how well an agent can manage infrastructure. Timeweb Cloud draws conclusions from this picture for the next few years: the programmatic layer that assistants access is becoming more important than the graphical interface. It is no longer enough for a provider to build a convenient panel;
Delegate Deep Research, Coding & Workflows to an AI Agent
Deploy autonomous multi-step execution with NomadNet Agent (agent.nomadnet.ai) — automated web research, data synthesis, code generation, and complex task orchestration.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.