sequence length

Sequence length or context length or context window is the maximum number of tokens that can be input into a transformer.

Here are a few examples:

Model	Sequence Length	Reference
GPT-4o	128K
Llama-3.1 405b	128K
DeepSeek-R1	128K
GPT-4.1	1M	1
Gemini 2.5 Pro	1M	2

Footnotes

1. Introducing GPT-4.1 in the API ↩

2. Gemini models. Gemini 2.5 Pro actually has a sequence length of $2^{20}=1,048,576$ tokens. ↩