Team
A hosted workspace for one team and its agents.
Talk to us
Talk to us- Hosted and updated for you
- All data sources and agent connections
- Datasets with per-dataset permissions
- Passkeys and two-step sign-in
- API keys for agents and automations
Pricing
Hosted for you, or run it yourself.
Every plan includes the whole product: all data sources, all agent connections, and the same memory for people and agents. Tell us about your team and we will send a quote.
Self-hosted
The full stack on infrastructure you control.
Talk to us
Talk to us- Everything in Team
- Runs on your own servers or cloud account
- Your data and your LLM key never leave your setup
- Postgres with pgvector and FalkorDB, in containers
- Invite-only registration
Enterprise
For organisations with their own requirements.
Custom
Talk to us- Everything in Self-hosted
- Help with deployment and migration
- Custom connectors and graph models
- Support with agreed response times
Questions
Before you ask.
- Which LLM does trellismem use?
- The one you configure. trellismem calls a language model to extract the graph and to write answers, and supports OpenAI, Anthropic, Gemini, Azure OpenAI, AWS Bedrock, Ollama and any OpenAI-compatible endpoint. Model usage is billed by that provider.
- Where is our data stored?
- In the databases of your deployment: Postgres for accounts, permissions and vectors, FalkorDB for the graph, and a storage volume for uploaded files. On a self-hosted plan all three run on your infrastructure.
- Can agents see everything?
- No. An agent acts with the API key it was given, and a key only reaches the datasets its owner can read.
- Can we remove data again?
- Yes. Deleting a document or a dataset removes its facts from the graph and from search.