# SysBench leaderboard

Instruction Following 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://modelchoice.dev/leaderboards/benchmarks/instruction-following-sys-bench

SysBench: system message following. Every model the catalog carries a published SysBench value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Instruction Following |
| Measure | Instruction success rate |
| Tasks | System-message constraints over conversations |
| Difficulty | Keeping long prompt rules |
| Published by | [SysBench: Can Large Language Models Follow System Messages?](https://github.com/PKU-Baichuan-MLSystemLab/SysBench) |

Tests how well a model follows the rules in a system message across single and multi-turn conversations.

## SysBench ranking

2 models with a published SysBench 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 | Instruction success rate | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Nemotron 3.5 Bolt E4B BF16](https://modelchoice.dev/models/nemotron-3-5-bolt-e4b-bf16) | NVIDIA | 67.88 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Nemotron 3.5 Bolt E4B](https://modelchoice.dev/models/nemotron-3-5-bolt-e4b) | NVIDIA | 67.22 | 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.
