Open WebUI

A feature-rich, self-hosted AI interface via Featherless.

Open WebUI

A feature-rich, self-hosted AI interface — connected to 19,000+ models via Featherless.

Featherless Managed Open WebUI is a fully managed deployment of Open WebUI with sandboxed compute and Featherless inference bundled into a single monthly subscription. You get a secure, always-on Open WebUI instance already pointed at the Featherless API. No Docker setup, no server management, no separate token bill.

Open WebUI handles the interface layer: chat, documents, voice, image generation, RAG, and multi-model support.

Featherless handles the model layer: 30,000+ open-source models, no prompt logging, flat-rate pricing.


Deploy

Go to featherless.ai/account/agents and click Browse Marketplace. Find Open WebUI and click Launch. You'll land on a Launch Open WebUI screen with two fields:

  • Name (optional): a friendly label for this instance, e.g. "My WebUI"

  • Model: a searchable dropdown

The default model is MiniMaxAI/MiniMax-M2.5 Not sure what to pick? Here are the available models on your current plan:

Model

Good for

NousResearch/Hermes-3-Llama-3.1-8B

General-purpose default, strong instruction following

LoganResearch/Nous-Hermes-ReflexAgent-8B-v1

Agentic tasks, self-reflection loops

MiniMaxAI/MiniMax-M2.5

Fast everyday assistant, balanced quality

cahlen/Convergent-7B

Lightweight, fast responses

arcee-ai/Trinity-Large-Thinking

Reasoning and thinking tasks

mims-harvard/TxAgent-T1-Llama-3.1-8B

Specialised task agent

ValiantLabs/Qwen3-1.7B-ShiningValiant3

Ultra-lightweight, low latency

ValiantLabs/Qwen3-4B-ShiningValiant3

Compact, fast general use

Featherlabs/Aura-7b

Featherlabs curated general assistant

rubra-ai/Meta-Llama-3-70B-Instruct

High quality, large instruction model

M-Alkassem/qwen2.5-coder-3b-final-merged

Code tasks, lightweight

The following models require a plan upgrade:

Model

Good for

moonshotai/Kimi-K2.5

Reasoning-heavy tasks, long context

inclusionAI/Ling-1T

Demanding tasks, massive MoE

Qwen/Qwen3.5-397B-A17B

Complex tasks, large MoE model

zai-org/GLM-5

Fast everyday assistant, strong multilingual

zai-org/GLM-4.7

Efficient mid-size assistant

Hit Launch. Your sandbox spins up in seconds and Open WebUI opens automatically at:

https://[port]-sbx-[your-id].sandbox.featherless.ai


Connect Featherless

Your instance comes pre-configured with Featherless as the model provider. If you ever need to re-add or verify the connection, go to Admin Settings → Connections → OpenAI-Compatible and confirm:

Click Save. Your models load automatically in the model dropdown at the top of the chat interface.


First conversation

In the chat interface, select a model from the dropdown at the top and send a message:

"Hey, I'm [your name]. I work in [field]. I prefer concise answers."

Try these while you're here:

Web search

"What are the most downloaded open-source LLMs on HuggingFace this week?"

Document analysis

Click the + icon in the chat input to upload a PDF or document. Then ask:

"Summarise the key points from this document."

Switch models mid-conversation

Click the model name at the top of the chat at any time to swap to a different model without losing your conversation.


Interface overview

Chat — your main conversation interface. Supports multi-turn conversations, file uploads, image input, and web search depending on the model.

Models — browse and manage all available models from your connected providers. Set a default model, pin favourites, and configure per-model system prompts.

Documents — upload files to create a knowledge base for RAG (Retrieval-Augmented Generation). Supports PDF, DOCX, TXT, and more.

Workspace — manage system prompts, model presets, and custom personas. Save configurations you use frequently.

Admin Settings — manage connections, users, and instance configuration. This is where you add or update your Featherless API connection.


Personalise your instance

System prompt

Set a default system prompt to give your assistant a consistent personality and set of instructions across all conversations.

Go to Workspace → Prompts and create a new prompt, or set a default in Admin Settings → Interface.

You are a helpful assistant powered by Featherless inference. You are direct, concise, and useful. You don't pad responses with filler. If you're uncertain, say so.

Model presets

Go to Workspace → Models to create custom model configurations — combine a base model with a system prompt, a temperature setting, and a name. These appear as selectable options in the chat dropdown.

Documents and RAG

Go to Documents to upload reference files. Once uploaded, reference them in chat with #document-name to pull relevant context into your conversation automatically.


Config reference

Open WebUI is configured via environment variables set at launch or via the Admin Settings panel in the browser interface.

Key environment variables

Variable

Description

OPENAI_API_BASE_URL

Your provider base URL — set to https://api.featherless.ai/v1

OPENAI_API_KEY

Your Featherless API key

DEFAULT_MODELS

Default model pre-selected in the chat dropdown

WEBUI_NAME

Custom name for your Open WebUI instance

ENABLE_RAG_WEB_SEARCH

Enable web search tool in chat (true/false)

These are pre-configured in your Featherless sandbox. You do not need to set them manually.


Troubleshooting

No models appearing in the dropdown. Go to Admin Settings → Connections and verify the Featherless base URL and API key are correct. Click the refresh icon next to the connection to reload the model list.

Model returns an error or empty response. The selected model may require an upgraded plan. Check for an "Upgrade to unlock" indicator on the model and switch to a free-tier model to confirm.

Document upload not working. Supported formats are PDF, DOCX, TXT, MD, and CSV. Files over 50MB may time out — split large documents into smaller chunks.

Voice input not responding. Microphone access must be granted in your browser. Open WebUI supports Whisper (local), OpenAI, Deepgram, and Azure STT — check Admin Settings → Audio to confirm your provider is configured.

Instance loads but shows a blank screen. Hard refresh the page (Ctrl+Shift+R or Cmd+Shift+R). If the issue persists, check the sandbox health indicator in your Featherless dashboard.


Resources

Last edited: Apr 15, 2026