Top news
Investing in the public benefit OfficialAnthropic NewsOct 18Popular Musk Pledges SpaceX Shares to ‘Every Kid in America’ WebBloomberg TechnologyOct 19Popular Websites are closing their doors to AI agents WebThe Verge AIOct 17 This is the new ChatGPT Voice, powered by GPT Live OfficialOpenAI on YouTubeOct 18 How AI giants will reshape global philanthropy WebAxios AIOct 19 Gemini 4 Argon Really Surprised Me WebTwo Minute PapersOct 19 Google Gemini 4 agents just beat Minecraft with no human intervention WebEngadget AIOct 19 Who could it be? 🧐 OfficialGoogle on YouTubeOct 18 Anthropic gives outside AI watchdog “employee-like” access WebArs Technica AIOct 18 What are self-replicating ‘AI worms’, and why is Anthropic worried about them? WebThe Guardian AIOct 18
GlossaryModelsNew to AI

Context window

Also written: Token context window, context length, context size, 100K context windows, Context anxiety, Context window bloat, Context overload, Context injection

The amount of text an AI model can consider at one time when processing input and generating output.

In plain terms

Think of the context window as the AI's short-term memory. It's the limit on how much information the AI can "see" and use from your conversation or the documents you provide at any single moment.

If you give the AI a very long document or have a long conversation, the AI might "forget" the beginning of the text or conversation once it reaches its context window limit. This means it might not be able to refer back to earlier parts when generating a response.

For example, if you ask an AI to summarize a book, but the book is longer than its context window, it can only consider a portion of the book at a time. To summarize the whole book, it would need to process it in chunks or have a larger context window.

The expert version

The context window defines the maximum number of tokens (words or sub-word units) that a large language model can process simultaneously as input and use to generate output. It represents the model's attention span for a given task.

Models with larger context windows can handle more extensive documents, maintain coherence over longer conversations, and perform complex reasoning tasks that require understanding relationships across large amounts of text. Tradeoffs include increased computational cost and memory requirements during inference for larger windows.

Why it matters now

Recent advancements from Google Gemini and OpenAI highlight significantly larger context windows, with models like Gemini 4 Argon boasting a million tokens. This expansion allows AI to process and reason over much longer texts, enabling more complex tasks and deeper analysis of documents and code.

Where it shows up 70 items, newest first
How each company uses it
Anthropic Claude
Anthropic's research explores methods for personalized AI to effectively use a reduced portion of conversation history, suggesting efficient context management.
Google Gemini
Google Gemini is pushing the boundaries with models like Gemini 4 Argon and Gemini 2.5 Pro, offering massive context windows of up to one million tokens for advanced text processing and reasoning.
OpenAI ChatGPT
OpenAI is developing new AI models that feature massive context windows, expanding the amount of information their systems can consider for complex tasks.
Related terms

TokensLarge Language ModelsAttention MechanismInferenceAI AgentsPrompt Engineering

Often confused with
Training Data
The context window is about what the model can process *now*, not
Memory
While similar, the context window is a specific technical limit for a si
Knowledge Cutoff
This refers to the date up to which the model's training data
Origin

The concept of a context window emerged with the development of transformer architectures, particularly in research on sequence-to-sequence models and attention mechanisms. It's a fundamental parameter in how these models process sequential data. Source ↗

Try this about 5 minutes

Paste a long article or a lengthy conversation into a chat interface with an AI like ChatGPT or Gemini. Then, ask the AI a question that requires it to recall information from the very beginning of the text or conversation. Observe if it can accurately reference the early details, demonstrating its context window in action.

Try it: Playbook entries that use thisPlaybook →

The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.

Questions people ask

How big is a context window?
Context windows are measured in tokens, which are roughly words or parts of words. They can range from a few thousand tokens to over a million tokens in the latest models.
Can I increase the context window?
You cannot directly increase the context window of a pre-trained model. However, developers can choose models with larger context windows or use techniques to manage information within the existing limit.

First seen on Subvolts Aug 28, 2023. Updated Oct 17, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.