# RULER (64K) leaderboard

Reasoning benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/reasoning-ruler64k

RULER at a 64K-token context. Every model the catalog carries a published RULER (64K) value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Reasoning |
| Measure | Overall accuracy at 64K |
| Tasks | Synthetic long-context tasks |
| Difficulty | Long-context retrieval and reasoning |
| Published by | [RULER: What's the Real Context Size of Your Long-Context Language Models?](https://github.com/NVIDIA/RULER) |

Synthetic long-context benchmark of retrieval, multi-hop tracing, aggregation and question answering, scored at a 64K-token context.

## RULER (64K) ranking

2 models with a published RULER (64K) value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Overall accuracy at 64K | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Nemotron 3.5 Bolt E4B BF16](https://modelscale.dev/models/nemotron-3-5-bolt-e4b-bf16) | NVIDIA | 86.99 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Nemotron 3.5 Bolt E4B](https://modelscale.dev/models/nemotron-3-5-bolt-e4b) | NVIDIA | 85.76 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
