Skip to content
ShopAssistShopAssist home

Use any OpenAI-compatible endpoint

OpenRouter, Groq, Together, Ollama, vLLM, LM Studio or your own server. If it speaks the OpenAI API with tool calling, it can run your shop assistant.

Presets

Requirements

  • Tool calling must work. For vLLM, start the server with auto tool choice and a tool-call parser. In LM Studio, use models with a tool chat template.
  • A context window of at least 16,384 tokens.
  • Up to three extra headers for access proxies.

Custom URLs are checked on every request: HTTPS only, public addresses only.

Supported models

Any model your endpoint serves that supports tool calling.

Where your data is processed

What your own AI key costs

Estimate based on list prices on 2026-09-28; your provider bills you directly.

Estimated provider cost per 1,000 conversations
ProviderModelUSD per 1,000 conversations≈ EUR per 1,000 conversations
Azure OpenAIGPT-6 Luna$3.61€3.10
MistralMistral Small 4$4.71€4.05
OpenAIGPT-5.6 Luna$7.62€6.55
Google GeminiGemini 3.1 Flash-Lite$8.85€7.61
OpenAIGPT-5 mini$9.85€8.47
Google GeminiGemini 3.5 Flash-Lite$12.02€10.34
MistralMistral Large 3$14.70€12.64
OpenAIGPT-5.4 mini$26.55€22.83
AnthropicClaude Haiku 4.5$36.10€31.05
AnthropicClaude Sonnet 5$72.20€62.09

Step-by-step setup in the docs

Frequently asked questions

Can I use a self-hosted model?

Yes, if it is reachable over public HTTPS with authentication and supports tool calling. Private and local network addresses are refused.

What context window do I need?

At least 16,384 tokens. For Ollama, set PARAMETER num_ctx 16384 or higher in the Modelfile.

Why is there a warning for routers?

Services such as OpenRouter, Together and Groq may route requests to other providers in other countries. Check their data-processing terms.

Try it on your store.

Free with your own AI key.