Verified65% confidenceFactExact time
The context windows of current mainstream large language models typically range from 8K to 200K tokens.
3
Sources
65%
Confidence
Medium-term (~90 days)
Relevance
9/21/2026
First Seen
Valid until: 12/20/2026
Sources
Related Entities
Related Claims
Unverified早期模型的上下文窗口仅有4K至8K个Token，约3000至6000个英文单词74% similarVerifiedThe context windows of current mainstream large language models have expanded to 128K or even longer.73% similarVerifiedThe context window of large language models has expanded from the early 4K to the million-token level.69% similarVerifiedThe context window of large language models is limited, and the length of text that can be processed in a single inference is typically tens of thousands to hundreds of thousands of tokens.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/938528API
curl https://kongchang.com/api/v1/knowledge/claims/938528MCP
get_claim(id=938528)