Skip to main content
By default, FlexOrch uses a locally hosted LLM (Ollama) for extraction tasks that deterministic pattern matching can’t fully resolve. If you’re on an Enterprise plan, you can bring your own API key to use a specific model or provider — giving you higher accuracy, access to the latest model versions, and full control over the LLM powering your pipelines.
BYO LLM is an Enterprise plan feature. To upgrade, contact sales@flexorch.com.

​
Supported providers

ProviderSupported models
OpenAIGPT-4o, GPT-4-turbo, GPT-3.5-turbo
Anthropicclaude-sonnet-4-6, claude-haiku-4-5
Azure OpenAIAny model you have deployed via your Azure endpoint
CustomAny OpenAI-compatible endpoint

​
Configure via the platform

1

Open LLM Configuration

In the FlexOrch dashboard, go to Settings → Integrations → LLM Configuration.
2

Enter your provider details

Select your provider from the dropdown, then enter the model name and your API key.
3

Test the connection

Click Test to verify FlexOrch can reach your provider with the credentials supplied.
4

Save

Click Save. FlexOrch will use your key for all subsequent pipeline runs that require an LLM.

​
Configure via API

You can also set your LLM configuration programmatically. This is useful for infrastructure-as-code setups or CI/CD pipelines that provision FlexOrch environments. 
curl -X POST https://api.flexorch.com/v1/settings/llm-config \
  -H "X-API-KEY: dfx_your_key_here" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "openai",
    "model": "gpt-4o",
    "api_key": "sk-..."
  }'

​
Request body

​
provider
string
required
The LLM provider to use. One of openai, anthropic, azure_openai, or custom.
​
model
string
required
The model identifier. For Azure OpenAI, use your deployment name. For custom endpoints, use the model name your endpoint expects.
​
api_key
string
required
Your API key for the selected provider. Stored encrypted at rest and never exposed in API responses.
​
endpoint
string
Required only for azure_openai and custom providers. The full base URL of the endpoint.
FlexOrch always attempts deterministic, rule-based extraction first. The LLM is only invoked as a fallback when pattern matching is insufficient. This means BYO LLM has no effect on documents where deterministic extraction succeeds — you only consume LLM tokens when it’s genuinely needed.