GenAiHub
Instruction Following & Long Context
8-needle retrieval tasks · 2026

MRCR v2 128K-256K

MRCR v2 slice focused on very long contexts at 128K-256K lengths.

Metric

Sequence match

Results shown

2

Publisher

BenchLM exact-source leaderboard

Snapshot

Fetched 2026-10-08T23:19:28.228Z

2 of 2 models

2 of 2 models

Higher is better

GPT-5.5 logo
GPT-5.5
Claude Opus 4.7 (Adaptive) logo
Claude Opus 4.7 (Adaptive)
List bars visualize relative performance. Plot bars use published 0–100 scores, shrink to a 1.5rem minimum, then scroll horizontally.

Scores are published snapshots and are only comparable within the same benchmark, version, metric, and evaluation configuration.