Which features use AI
Semantic search is worth calling out because it looks like an AI feature and is often assumed to be one. It runs on a local embedding service (
all-MiniLM-L6-v2) that ships as part of the deployment, so playbook content is never sent to an external provider in order to be searchable. Search works whether or not AI is enabled.
Who provides the model
You do — in both deployment models. Cymph does not bundle, resell or operate a model on your behalf. An administrator enters a base URL, an API key and a model name, and Cymph calls that endpoint using an OpenAI-compatible API. See AI configuration. Because the base URL is yours to choose, you can point Cymph at a commercial API, at a provider in a specific jurisdiction, or at a model hosted inside your own network. If you need AI processing to stay in a particular region for GDPR or data-residency reasons, you satisfy that by choosing a provider that meets the requirement — it is not a property of the Cymph deployment. This has a direct consequence for your data-processing analysis: the relationship for AI features is between you and the provider you select. Cymph is not a processor in that path and cannot see the requests. This holds for a managed cloud tenant exactly as it does for a self-hosted deployment — the AI provider is a customer-configured setting in both.What reaches the model
Only the content needed for the feature you invoked:- Playbook generation — the prompt you write, and the context of the playbook being generated.
- Mindmap generation — the prompt or source material you supply.
- Autotagging — the playbook content being tagged.
- Playbook translation — the title and description of the playbook and of its nodes. Other fields are not sent.
- Document-to-workflow conversion — the playbook’s documentation text. Workflow-to-document conversion — the workflow’s steps, their properties and their connections. In each direction only the side being converted from is sent.
- Cymph AI assistant — the conversation, any file you attach to it, and the workspace content the assistant retrieves to answer: the playbooks, assets, framework mappings and execution records you have access to. The assistant cannot retrieve content that is not shared with you.
Provider credentials
The AI provider’s API key is stored in the deployment’s database encrypted with that deployment’s own key material, and decrypted only in memory when a request is made. In a self-hosted deployment it never leaves your infrastructure. See Data protection.Cost
Cymph charges no per-token, per-request or per-seat fee for AI features. Your licence or plan covers the AI features themselves; the model calls they make are billed by the provider you configure, under your own account and API key, at that provider’s rates. A model hosted inside your own network has no per-use charge at all. Semantic search runs on the bundled embedding service and has no usage cost in either deployment model.Model versions and updates
Three components are involved, and each is updated on a different path.
Chat model. The model name you enter is passed verbatim to your provider on every request. Cymph never substitutes, upgrades or changes it. If your provider offers pinned identifiers (for example a dated model version), use one and the model behind the AI features stays fixed until you change the setting. When a provider retires a model, requests start failing and the Test button on the AI configuration page reports the error; entering the replacement model name resolves it.
Embedding model.
all-MiniLM-L6-v2 ships inside the deployment and is not updated independently of the product. If a future release replaces it, that is a product change called out with the release, and the embeddings are regenerated inside the deployment as part of the upgrade — no content leaves the deployment to do so.
Prompts and feature behaviour. The prompts Cymph sends and the way each AI feature works are part of the product and change with product releases. In a self-hosted deployment you decide when to upgrade; in a managed cloud tenant Cymph performs upgrades — see Deployment models.

