AI and agents · Rio de Janeiro or São Paulo

VPS with Qdrant, ready to use

Qdrant is a vector database that stores embeddings and finds the content most similar to a query. It is the base for semantic search and RAG. At Mega it comes preinstalled on a VPS with Debian 13, Docker and HTTPS, in Rio de Janeiro or São Paulo, Brazil.

Installed on Debian 13 with Docker
HTTPS and API key already configured
Data center in Rio de Janeiro or São Paulo, Brazil
24/7 support

qdrant.yourdomain.com

Collections in the Qdrant dashboard, with status, number of points, and

Points in a collection: each product with its vector and the data

When you open the dashboard, Qdrant asks for the API key before showing

Collections in the Qdrant dashboard, with status, number of points, and3 screens

Plans

VPS plans for hosting Qdrant

Qdrant keeps vectors in memory to respond quickly, so RAM determines how large your dataset can be. The app fits in as little as 512 MB of free memory, and the Basic plan is the starting point we recommend.

Location

Selected

Start

US$ 6.95/mo

1 vCore CPU
1 GB Memory
NVME 20 GB Disk
Unlimited Traffic

Ask on WhatsApp

Recommended

Basic

US$ 11.95/mo

2 vCore CPU
2 GB Memory
NVME 40 GB Disk
Unlimited Traffic

Ask on WhatsApp

Plus

US$ 16.95/mo

2 vCore CPU
4 GB Memory
NVME 80 GB Disk
Unlimited Traffic

Ask on WhatsApp

Pro

US$ 21.95/mo

4 vCore CPU
6 GB Memory
NVME 120 GB Disk
Unlimited Traffic

Ask on WhatsApp

Billed monthly in US dollars, with the app installed. Rio de Janeiro or São Paulo, same price. Plans start at the smallest size where the app runs well. Need more? Upgrade from the client area, without reinstalling.

Where the server lives

Qdrant in Rio de Janeiro

With Rio de Janeiro selected, your VPS with Qdrant runs in Mega's original data center, on our own network with a presence at the Rio internet exchange (IX.br) and redundant transit. It is the lowest latency for users in Rio, Espírito Santo and Minas Gerais, and our support team works in the same city as the server.

Measured latency, not a promisemeasured 18 minutes ago

Rio de Janeiro to

Vivo11.4ms

Claro19.5ms

TIM20.2ms

Oi11.1ms

Algar4.5ms

São Paulo to

Vivo15.0ms

Claro8.6ms

TIM23.7ms

Oi15.2ms

Algar7.0ms

Ping from real customer connections on each carrier to our data centers in Rio and São Paulo, measured every hour by independent probes (Globalping). Algar: measured from our edge, where we connect to it directly.

What is Qdrant

Qdrant is an open-source vector database written in Rust and licensed under Apache 2.0. An AI model turns each piece of text or product into a vector, and Qdrant returns the closest matches to a query in milliseconds.

In practice, it is the component that lets an AI assistant answer from your own documents, a technique called RAG. You split the text, generate the vectors and store them in Qdrant. For each question, it hands the model the most relevant passages.

You access it through a REST or gRPC API, with filters on each item's fields (payload) and a web dashboard for browsing collections and points. In the Mega installation, both the API and the dashboard only respond to whoever has the key generated for your server.

Source: Official Qdrant repository on GitHub, accessed in October 2026

What you get with the VPS

Qdrant 1.19.2 in a Docker container, with automatic restart after reboot
Debian 13 with Docker already installed
Your own HTTPS address, with a Let's Encrypt certificate
API key generated during installation, sent by email and WhatsApp
Qdrant telemetry disabled
Root access via SSH
VPS panel in the client area: restart, reinstall, change password and console

What a Qdrant server is used for

Chatbot over your documents

Manuals, contracts and a knowledge base feed an assistant that answers by citing the right passage.

Semantic search for a store or site

Someone who types "gift for a 5-year-old" finds related products, even if the descriptions don't use those words.

Memory for AI agents

The agent stores conversations and facts and retrieves whatever is useful when it needs it.

Item recommendations

Suggest products, articles or courses similar to what the customer has already viewed or bought.

n8n workflows

n8n has a Qdrant node, so you can index files and query the database inside your automation.

Finding duplicates

Spot near-identical records, tickets or texts that a regular search would miss.

A look inside Qdrant

Collections list in the Qdrant dashboard, showing status, number of points and vector configuration (example).

Points in a collection, each with its vector and associated data, the payload (example).

Dashboard home screen asking for the API key before showing any content (example).

Screenshots taken by the Mega Dedicados team on a test installation, with no customer data.

What comes configured on your Qdrant

Automatic HTTPS

Your own address with a Let's Encrypt certificate, with no proxy or renewal to manage.

API key access

Without the key, the API and the dashboard refuse the connection. The key is generated during installation and belongs only to you.

Automatic restart

The Docker container starts by itself if the server reboots.

REST and gRPC API

Ready for the official clients in Python, JavaScript, Go, Rust, Java and .NET.

Web dashboard

Browse collections and points and test searches right from your browser.

Telemetry off

Qdrant does not send usage data outside your server.

How to order

1
Choose the plan and location
Select Start, Basic, Plus or Pro and choose whether the server is in Rio de Janeiro or São Paulo, Brazil. The price is the same in both.
2
Make the payment
Pix, card or boleto, on a monthly, semiannual or annual cycle. No lock-in contract.
3
Wait for the installation
We create the VPS and install Qdrant with HTTPS automatically, in about 5 minutes.
4
Receive your access details
The address and API key arrive by email and WhatsApp, and you can create your first collection right away.

Which plan to choose for Qdrant

In our tests, idle Qdrant uses about 90 MB of memory. Real usage grows with the number of vectors and the size of each one.

Plan	Best for
Start 1 vCore · 1 GB	Prototypes, tests and small datasets, with a few thousand documents.
Basic 2 vCore · 2 GB	Recommended to get started: an AI assistant in production with a medium-sized dataset.
Plus 2 vCore · 4 GB	Qdrant and n8n on the same server, building a complete RAG workflow.
Pro 4 vCore · 6 GB	Large datasets with millions of vectors, or several apps sharing the VPS.

Frequently asked questions

FAQ about VPS with Qdrant

What is Qdrant and what is it used for?

It is an open-source vector database. It stores embeddings and returns the items most similar to a query, which enables semantic search, RAG, recommendations and memory for AI agents.

How much does Qdrant hosting cost?

Qdrant is free, so you only pay for the VPS. The Start plan costs US$ 6.95 per month, and Basic, which we recommend, costs US$ 11.95. Payment is by Pix, card or boleto, with no lock-in and a 7-day refund period.

What is the minimum VPS for running Qdrant?

The app fits in as little as 512 MB of free memory, and idle Qdrant uses about 90 MB. For real use, the Basic plan (2 vCore, 2 GB RAM and 40 GB NVMe) is the safest starting point.

Is self-hosted Qdrant worth it?

If you want your data on a server you control, in a data center in Brazil and at a fixed cost, yes. You get root access, choose when to update and don't pay per query, since traffic is unlimited.

Does Qdrant create embeddings by itself?

No. Vectors come from an AI model, such as OpenAI's or a local model. Qdrant stores and searches them, and tools like n8n and LangChain connect the model to the database.

Is my vector database exposed to the internet?

Not without the key. The API and the dashboard require the key generated during installation, which only the customer receives, and the connection uses HTTPS.

Can I run Qdrant together with n8n or other apps?

Yes. Each app gets its own address on the same VPS, as long as the memory is enough for all of them. For Qdrant and n8n together, the Plus plan is usually the most comfortable.

How do I update Qdrant and back it up?

The installed version is 1.19.2 and it only changes when you decide, via SSH with root access. For backups, use Qdrant snapshots, which copy an entire collection and let you move it to another server. Mega's 24/7 support covers the server side, while Qdrant usage is documented by the project.

Updated on October 19, 2026