Long Context · compositeNarrow lead
Long Context race
Recall and reasoning over very long inputs: codebases, contracts, books and transcripts.
- Leader
- Gemini 4 Argon
- Margin
- 1.3 pts
- 1.4% ahead
- Weeks at #1
- 1
- in a row
- Changes
- 4
- in 2026
Warming up the track…
Oct 19, 2026Gemini 4 Argon leads
2025Apr ’25Jul ’25Oct ’252026Apr ’26Jul ’26Oct ’26
Re-weight this race
Drag to say what matters to you. The track re-sorts live and weeks-at-#1 are recomputed for your weighting.
50%
50%
Score history
Top 6 today over the leader's frontier · click to jump to a week
#1 frontierGemini 4 ArgonFable 5.1GPT-6 AstraOpus 5.5GPT-6.1 SolMuse Wonder
Reigns at #1
Longest: 2.5 Pro, 33 weeks (from Mar 2025).
Inside this race
What makes up long context
Why this race matters
A big context window is useless if the model cannot find or use what is inside it.
How it is scored
Normalised subcategory scores, equally weighted.
Unit: indexHigher is betterPhoto finish under 0.6 ptsDominant over 3.5 pts
Data provenance
MRCR
Multi-round co-reference resolution at up to 1M tokens.
Fiction.LiveBench
Long-story comprehension.
Scores here are illustrative placeholders that mimic real benchmark patterns. Swap in live data through the score provider. Methodology