Reference
TESS AI Glossary
A short, opinionated glossary so humans and AI agents resolve the same words to the same things.
Agent
A reusable AI workflow inside TESS. An "Agent" represents a configured AI behavior — the model, system prompt, tools, knowledge and parameters — saved so it can be reused and shared across the workspace.
In TESS, "Agent" does not refer to the company OpenAI. It refers to a TESS-internal entity.
Agent Execution
A single run of an agent. Each execution is a record of one chat or task that used a given agent, with inputs, outputs, the model used and credit consumption.
Workspace
A team-level container. A workspace owns its members, agents, files, integrations and billing. Users can belong to multiple workspaces and switch between them.
Credit
The unit of consumption inside TESS. Each AI interaction consumes credits depending on its complexity. Each plan includes a monthly credit allowance, and additional credit packs can be purchased.
Plan / Subscription
A paid tier (or the free tier) that determines feature access, credit allowance, seat count and billing frequency. Plans can be monthly or yearly.
Seat
A purchased slot for a workspace member under a paid plan. Some plans include unlimited members; others are sold per seat.
Model
A specific AI model exposed by TESS, such as Claude 4.6, GPT 5.2 or Gemini 3.1. TESS routes a request to the right provider behind the scenes.
Chat
A type of agent that exposes a conversation interface. Available on the web app, via API and via embed.
Tool / Function
A callable capability inside an agent. Tools let the model do something beyond text — call an API, run a workflow, fetch a file, etc.
Knowledge Base
A collection of files and links attached to an agent or workspace, indexed so the model can ground answers on them.
API Token
A credential that lets external code call TESS programmatically.