Open WebUI in Russia: connect Claude, GPT, Gemini and DeepSeek without a VPN or a foreign card
Open WebUI is a web chat interface for AI models that you install on your own computer or server, often together with Ollama. Besides local models it works with any OpenAI-compatible API: RuAPI adds Claude, GPT, Gemini, DeepSeek and Qwen to the same menu with one key. Requests go to www.ruapi.ai without a VPN, and the balance is topped up in USDT — no foreign card needed.
from openai import OpenAI
client = OpenAI(
api_key="sk-...",
base_url="https://www.ruapi.ai/v1",
)
resp = client.chat.completions.create(
model="deepseek-flash",
messages=[{"role": "user", "content": "Hello"}],
)
print(resp.choices[0].message.content)Use this example (cURL tab) to test your key in a terminal. In Open WebUI, enter the same address https://www.ruapi.ai/v1 in the URL field and the same key in the API Key field of an OpenAI connection.
Why connect RuAPI to Open WebUI
Claude, GPT and Gemini do not run locally.
Through Ollama, Open WebUI runs open-weight models — Llama, Qwen, Gemma, DeepSeek-R1 and others (ollama.com/library, 1 Nov 2026). The weights of Claude, Gemini and the GPT-5 and GPT-6 series are not published, and large open models need a lot of memory and a powerful GPU. Through RuAPI these models run in the cloud, while Open WebUI stays with you.
OpenAI and Anthropic keys are out of reach from Russia.
Russia is not on the supported-country lists of Anthropic and OpenAI (anthropic.com/supported-countries, developers.openai.com, 1 Nov 2026), and Russian bank cards do not work abroad. A RuAPI account is created with email, Google or Yandex, and the balance is topped up in USDT from 5 USDT.
One key instead of a connection per vendor.
To have Claude, GPT and Gemini in Open WebUI, people usually set up a separate connection and key with each company. With RuAPI it is one OpenAI connection: models from different vendors in one model menu, and the cost of every request in the console log.
How to connect RuAPI to Open WebUI: 3 steps
You need a running Open WebUI (Docker, pip or Open WebUI Desktop — see the FAQ below) and admin rights: the first user to sign up becomes the administrator.
Sign up for RuAPI and top up
Sign up at www.ruapi.ai with email, Google or Yandex, no phone number. Top up in USDT from 5 USDT (TRC20, BEP20, TON, Polygon); USDT can be bought with rubles via SBP — guide: docs.ruapi.ai/en/buy-usdt. Foreign-issued cards are accepted too.
New accounts usually get $0.05 right away: on the trial balance claude-haiku-5-5, deepseek-flash and gpt-6-luna work. The other models in the list unlock after your first top-up.
Create a key for Open WebUI
Console → API Keys → "Create API key". If you open the models to colleagues or family, give the key a spending limit: by default only the admin sees the models of a new connection, and once they are made Public in Admin Panel → Settings → Models, the chats of all users spend one RuAPI balance.
Add an OpenAI connection in Open WebUI
- Open Admin Panel → Settings → Connections and, under Manage OpenAI API Connections, click "+" (Add Connection).
- URL: https://www.ruapi.ai/v1 — with /v1. API Key: your sk-… key. The icon next to URL (tooltip Verify Connection) checks the address and the key.
- Leave API Type on Chat Completions (the default): in Responses mode RuAPI passes GPT requests through as is and converts them for other models, but Chat Completions is the safer choice for Open WebUI. To keep only the models you need, expand Advanced and add them one by one under Model IDs (type the ID → "+"), for example deepseek-flash, claude-sonnet-5-5 and gpt-5.4. If you leave it empty, the whole RuAPI catalog ends up in the menu, including image and video models. Click Save, open a new chat and pick the model in the list above the chat.
Full guide with troubleshooting: docs.ruapi.ai/en/integrations/open-webui
What to set in Open WebUI
OpenAI connection
URL: https://www.ruapi.ai/v1
API Key: sk-…
Docker variables
OPENAI_API_BASE_URL=https://www.ruapi.ai/v1
OPENAI_API_KEY=sk-…
Payment
USDT / USDC from 5 USDT
foreign-issued cards
Cloud model prices for Open WebUI
What the models worth keeping in Open WebUI next to your local ones cost, in USD per 1M tokens, from the pricing page. No subscription: only the tokens spent in your chats are charged.
| model | input $/M | output $/M |
|---|---|---|
| deepseek-flash | $0.17 cache read $0.0120 | $1.44 |
| qwen3.8-flash | $0.14 cache read $0.0288 | $0.49 |
| gpt-5.4-mini | $0.90 cache read $0.0900 | $5.40 |
| claude-haiku-5-5 | $0.12 cache read $0.0120 >100K: $0.60 | $0.60 >100K: $3.00 |
| gemini-3.5-flash | $1.80 cache read $0.18 | $10.80 |
| gpt-5.4 | $3.00 cache read $0.30 >272K: $6.00 | $18.00 >272K: $27.00 |
| claude-sonnet-5-5 | $2.40 cache read $0.12 | $12.00 |
Columns: input is new input tokens, “cache read” is input tokens served from the cache, output is the model’s reply. With every new message the chat sends the whole previous conversation to the model, so long chats cost more than short ones. For GPT and DeepSeek the repeated start of a request is read from the cache automatically; for Claude in the OpenAI format, by cache_control marks, and if the client sets none, RuAPI adds them itself when the next message comes within 5 minutes (docs.ruapi.ai/en/prompt-caching).
The grey “>100K” line is the vendor’s rate for long prompts: if a prompt goes over the threshold, the whole request is billed at that rate.
Local Ollama or cloud models through RuAPI
| Open WebUI + RuAPI | Open WebUI + local Ollama | |
|---|---|---|
| Where the model runs | In the cloud; the request goes through www.ruapi.ai to the model provider | On your computer or server; data stays there |
| Hardware | Any machine that can run Open WebUI | Memory and a GPU sized for the model |
| Models | Claude, Gemini, GPT, plus DeepSeek, Qwen, GLM, Kimi | Open-weight models: Llama, Qwen, Gemma, DeepSeek-R1, gpt-oss and more |
| Cost | Tokens only: deepseek-flash is $0.168 per 1M input and $1.44 per 1M output | No token fees; the costs are hardware and electricity |
Ollama information is from ollama.com/library and github.com/ollama/ollama, checked 1 Nov 2026. Ollama also offers cloud models tagged cloud; this table compares running on your own hardware. Both can be combined in one Open WebUI.