What it is
A token is the fundamental unit of text processing for large language models. Models break text into tokens — subword chunks that roughly correspond to 3/4 of an English word. Every AI interaction has token limits: an input limit (how much context you can provide) and an output limit (how long the response can be).
Why it matters
Token limits directly affect what your AI agents can do. An agent with a 4K token context window cannot read a 50-page document in one shot. Understanding tokens helps you design prompts, configure grounding, and estimate costs.
Key components
- Input tokens
- Output tokens
- Context window
- Tokenization
- Cost per token
How it connects
Agentforce and Salesforce AI features consume tokens on every interaction. Token usage factors into Flex Credit consumption and affects response quality.
Good to know
More tokens does not mean better. Stuffing too much context into a prompt can actually degrade quality. Be selective about what data you ground your agents on.
Related terms
LLM (Large Language Model)
The AI technology behind ChatGPT, Claude, and the intelligence in Agentforce. Trained on massive amounts of text to understand and generate human language.
Prompt Engineering
The practice of crafting precise instructions to guide an AI model's behavior, capabilities, and limitations.
Flex Credits
Salesforce's consumption-based pricing unit for Agentforce — you pay per conversation, not per user license.