SRE Bench
As of Nov 3, 2026, SRE Bench (v1) is maintained by Vals AI. Metric accuracy (%). Ceiling 100. Saturation unknown. Contamination risk low.
- Maintainer
- Vals AI
- Metric
- accuracy (%)
- Ceiling
- 100
- Saturation
- unknown
- Contamination
- low
- Cadence
- Declared cadence: as Vals updates (checked daily). Catalog snapshots: 1, latest Oct 27, 2026.
- Source
- Vals AI · this benchmark on Vals AI
- Grading
- Automatic (tests or exact checks) — Reverse-engineering tasks with deterministically verifiable answers.
Source: Vals AI. A proprietary private test set run by Vals; scores are republished with permission. Rows are dated the day isitsota first saw them, and only the benchmark's standard harness counts.
Categories
Where it counts
- Cybersecurity — secondary, weight 0.5
Top 5
Current top scores
- GPT-6 Astra · 56.87 · Oct 27, 2026
- GPT-6.1 Sol · 50.763 · Oct 27, 2026
- Gemini 4 Argon · 44.275 · Oct 27, 2026
- Claude Opus 5.5 · 33.588 · Oct 27, 2026
- GPT-5.6 Sol · 30.534 · Oct 27, 2026
| Model | Score | Date | Kind |
|---|---|---|---|
| GPT-6 Astra | 56.87 | 2026-10-27 | snapshot |
| GPT-6.1 Sol | 50.763 | 2026-10-27 | snapshot |
| Gemini 4 Argon | 44.275 | 2026-10-27 | snapshot |
| Claude Opus 5.5 | 33.588 | 2026-10-27 | snapshot |
| GPT-5.6 Sol | 30.534 | 2026-10-27 | snapshot |
| GPT-6 Sol | 30.534 | 2026-10-27 | snapshot |
| Claude Sonnet 5.5 | 30.153 | 2026-10-27 | snapshot |
| Claude Fable 5.1 | 22.901 | 2026-10-27 | snapshot |
| Claude Opus 5 | 12.214 | 2026-10-27 | snapshot |
| Muse Wonder Full | 11.832 | 2026-10-27 | snapshot |
| Gemini 3.8 Flash | 11.069 | 2026-10-27 | snapshot |
| GPT-5.6 Terra | 9.542 | 2026-10-27 | snapshot |
| GLM 5.3 | 9.16 | 2026-10-27 | snapshot |
| Claude Haiku 5.5 | 5.725 | 2026-10-27 | snapshot |
| Gemini 3.7 Flash | 4.58 | 2026-10-27 | snapshot |
| Mimo V2.6 Flash | 4.198 | 2026-10-27 | snapshot |
| Grok 4.7 | 3.817 | 2026-10-27 | snapshot |
| GPT-5.5 | 3.817 | 2026-10-27 | snapshot |
| Muse Spark 1.3 | 3.817 | 2026-10-27 | snapshot |
| Mimo V2.6 Pro | 3.053 | 2026-10-27 | snapshot |
| GPT-6 Luna | 2.672 | 2026-10-27 | snapshot |
| Hy4 Preview | 2.29 | 2026-10-27 | snapshot |
| DeepSeek V4 Pro 0813 | 1.908 | 2026-10-27 | snapshot |
| GPT-5.6 Luna | 1.527 | 2026-10-27 | snapshot |
| DeepSeek V4.1 Flash | 0.763 | 2026-10-27 | snapshot |
| GLM 5.3 Flash | 0.763 | 2026-10-27 | snapshot |
| Grok 4.5 | 0.763 | 2026-10-27 | snapshot |
| Claude Sonnet 5 | 0.382 | 2026-10-27 | snapshot |
| Step 5 Preview | 0.382 | 2026-10-27 | snapshot |
| GLM 5.2 | 0 | 2026-10-27 | snapshot |