isitsota / categories / Long context
Long context · Oct 13, 2026
Who is SOTA in Long context?
Nobody yet.
Usable long context, not a needle-in-a-haystack demo. Fiction.liveBench at 120k tokens plus longer-query preference. As of Oct 13, 2026, Long context has no model with enough independent evidence for a SOTA crown.
No crown on this slice. Empty categories stay empty: embeddings, speech, and Terminal-Bench 4.0 are not invented from missing exports. SWE-bench Verified is display-only.
Standings · le-32b
The board
| # | Model | State | Confidence |
|---|
Evidence recipe
What can move this crown
- Fiction.liveBench 120k — primary, weight 1
- LMArena Text (style-controlled, longer query) — primary, weight 0.7