Context window
The amount of text (measured in units called tokens, roughly corresponding to word fragments) a large language model can consider at once when generating a response, including both the current prompt and any conversation history provided. A model with a larger context window can work with longer documents or longer conversations in a single exchange without losing track of earlier content.
Defined in 5 GAGE programs, which carry 9 distinct definitions of it. The wording above is taught in The AI Lobbyist: Certified AI Policy Strategist.
How each discipline defines it
The same term does different work depending on who is using it. These are the definitions as each program teaches them, unedited.
The amount of text (measured in units called tokens, roughly corresponding to word fragments) a large language model can consider at once when generating a response, including both the current prompt and any conversation history provided. A model with a larger context window can work with longer documents or longer conversations in a single exchange without losing track of earlier content.
The maximum amount of text a model can consider at once, covering the system prompt, the conversation history, any retrieved passages, and the user's question together. When the material you want to supply exceeds the window, something must be dropped or summarized, which is itself a configuration decision about what the model gets to see.
The portion of a conversation's text that a language model is actively given to read before generating its next response, which is how the tool appears to "remember" earlier details within a single ongoing conversation, even though it has no independent memory of that conversation once it ends.
The maximum amount of text an AI model can consider at once. Claude's standard context window is about 200,000 tokens (roughly 150,000 words); a 1,000,000-token window is available on newer Claude models for higher usage tiers. Gemini Pro models offer 1,000,000+ tokens.
the finite amount of text a model can consider at once when producing a single response, including the current conversation and anything retrieved into it; distinct from memory, which is the mechanism deciding what fills that window.
Where it is taught
The exact lessons this term appears in. The first module of every program is free with a free account.
- Building Your AI Briefcase · Your AI Learning Companion, Certified AI Practitioner: Workplace Foundations
- What is AI? Definitions, History, and Types · AI Fundamentals, Certified AI Practitioner: Workplace Foundations
- How AI Actually Works: Algorithms, Data, and Training · AI Fundamentals, Certified AI Practitioner: Workplace Foundations
- Multimodal AI: Text, Image, Audio, Video · AI Fundamentals, Certified AI Practitioner: Workplace Foundations
- The Eight Components I · The Agent, Deconstructed, Certified Agentic AI Governance Professional (CAAGP)
- Prompts, context, and why the same model gives different companies different answers · Build Before You Govern, Certified AI Governance Professional (CAIGP)
- What AI Really Is: No Hype, No Myths, No Fear · Foundations of AI for Lobbyists, The AI Lobbyist: Certified AI Policy Strategist
- How AI Actually Works Under the Hood: Architectures and Limits for Policy People · Technical Credibility Deep Dive, The AI Lobbyist: Certified AI Policy Strategist
- Watching a machine run the whole conversation · The Demystification Lab, Senior AI Scam Defense
Where it shows up at work
A definition is where Context window starts. The lessons above teach it inside the competencies below, which are what job descriptions ask for and what the GAGE credential records. Every link here is drawn from the same careers map that builds a role's readiness check.
Explains training, tokens, context windows, embeddings, retrieval and fine-tuning well enough to ask an engineer a precise question and spot weak evidence.
3 lessons on this page teach it as a primary outcome.
Builds production systems on foundation models: retrieval augmented generation, structured tool use, evaluation harnesses, guardrails, and cost and latency budgets.
2 lessons on this page teach it as a primary outcome.
Briefs a model like a good manager: context, constraints, examples, verification, and knows when to stop delegating.
One lesson on this page teach it as a primary outcome.
Roles that hire for it
Each role page shows the full competency set, what the seat decides, and a readiness check against your own evidence.
- AI Engineer, production systems editionneeds it at depth · 23 open postings this week
- Chief Technology Officer, AI engineering editionneeds it at depth
- AI Security Architectneeds it at depth · 8 open postings this week
- AI Evaluation Specialistneeds it at depth · 6 open postings this week
- AI Governance Engineerneeds it at depth · 7 open postings this week
- AI Model Validatorneeds it at depth
How job descriptions say it
Postings rarely use the textbook word. These are the phrases the careers map reads as asking for the competencies above.
Where the frameworks place it: NIST AI RMF.
One interview question about it
Explain how a large language model produces an answer, at the depth a governance decision needs and no deeper.
Asked of candidates for how models work, at a governance depth. Answering it in your own words is exactly how a GAGE topic is passed, and the credential records the pass where an employer can check it.
If you come from another field
Some backgrounds already carry this competency, and the readiness check gives credit for it.
- Data analysis, stewardship and business intelligence carries it as a claim to back up: SQL and analytical fluency make the technical layer approachable.
- Model risk, credit risk and quantitative analysis carries it as work you can show: Quantitative fluency is demonstrated in your work.
- A computer science or software engineering degree carries it as work you can show: You have trained or integrated a model in coursework or a project.
Terms it appears with
Not an alphabetical neighbourhood: these are the terms taught in the same lessons, ranked by how often they appear together.