NewHow we catch a seller listing the wrong model, and refund you if one does.

Open WebUI with MartLLM: every model from one connection

In Open WebUI go to Settings › Admin › Connections, add an OpenAI API connection with URL https://api.martllm.com/v1, Bearer auth and your MartLLM key, keep API Type on Chat Completions, verify and save. For a new Docker install, pass OPENAI_API_BASE_URL and OPENAI_API_KEY instead.

Steps (existing install)

  1. Make a key on your keys page and add credit.
  2. Sign in to Open WebUI as an admin and go to Settings › Admin › Connections.
  3. Under Manage OpenAI API Connections, click + Add Connection.
  4. Fill in:
    • Connection Type: External
    • URL: https://api.martllm.com/v1
    • Auth: Bearer
    • API Key: your ull_… key
    • API Type: Chat Completions
  5. Click Verify Connection, then Save.
  6. Open a new chat. The model picker now lists MartLLM's models.

Steps (new Docker install)

docker run -d -p 3000:8080 \
  -e OPENAI_API_BASE_URL=https://api.martllm.com/v1 \
  -e OPENAI_API_KEY=ull_your_key \
  -v open-webui:/app/backend/data \
  --name open-webui --restart always \
  ghcr.io/open-webui/open-webui:main

These variables only apply the first time Open WebUI starts. After that the value saved in its database wins, so on an install that already ran, use the screen above (or start it once with RESET_CONFIG_ON_START=true).

To keep your current OpenAI connection and add MartLLM next to it, use the plural variables with a semicolon between entries:

-e OPENAI_API_BASE_URLS="https://api.openai.com/v1;https://api.martllm.com/v1" \
-e OPENAI_API_KEYS="sk-your-openai-key;ull_your_key"

A tidy model list

MartLLM has hundreds of models. Two settings under Advanced in the connection keep the picker short:

  • Model IDs: list only the ids you want, for example deepseek/deepseek-v4.1-flash and google/gemini-3.5-flash. Open WebUI then shows just those.
  • Prefix ID: a word such as uberllm is added in front of every model name (uberllm.deepseek/deepseek-v4.1-flash), so you can tell them apart from other connections. It is removed before the request is sent.

Troubleshooting

  • Verify fails: check the URL ends in /v1 with nothing after it, and the key has no space.
  • Model list loads slowly or is empty: Open WebUI gives the list 10 seconds by default. Add the ids under Model IDs so it stops fetching the full list, or raise AIOHTTP_CLIENT_TIMEOUT_MODEL_LIST.
  • Changed the Docker variables and nothing happened: see the first-start note above.
  • 402 insufficient credits: add credit on the credits page.
  • Text arrives in bursts behind your own nginx: turn off proxy buffering for Open WebUI.
  • Documents and web search: Open WebUI makes embeddings with its own local model by default, which is what you want; MartLLM does not serve embeddings.

Keep reading

Models worth pinning in Open WebUI

Good everyday models. Add their ids under Model IDs to show only these.

ModelModel id (paste this)Price now
DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash$0.0075 in · $0.0282 out / 1M
Gemini 3.5 Flashgoogle/gemini-3.5-flash$0.0075 in · $0.0450 out / 1M
Claude Sonnet 5.5anthropic/claude-sonnet-5.5$0.3600 in · $1.80 out / 1M
GPT-5.5openai/gpt-5.5$1.25 in · $7.50 out / 1M
Kimi K3moonshotai/kimi-k3$0.0384 in · $0.2021 out / 1M
GLM 5.3z-ai/glm-5.3$0.0029 in · $0.0100 out / 1M
Gemma 3 27Bgoogle/gemma-3-27b-it$0.00079 in · $0.0020 out / 1M
Qwen3 VL 235B A22Bqwen/qwen3-vl-235b-a22b-instruct$0.0021 in · $0.0188 out / 1M

Prices at the last hourly build, no fee added · every model →

FAQ

How do I add MartLLM to Open WebUI?

Settings › Admin › Connections › + Add Connection: URL https://api.martllm.com/v1, Auth Bearer, your MartLLM key, API Type Chat Completions. Verify, save, and the models appear in the picker.

Why did my OPENAI_API_BASE_URL change not apply?

Open WebUI reads those variables only on first start and then keeps the saved value. Change the connection in the admin screen, or start once with RESET_CONFIG_ON_START=true.

Can Open WebUI use MartLLM for embeddings?

No, MartLLM does not serve embeddings. Keep Open WebUI on its default local embedding model for documents.