# GPQA Diamond 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://modelchoice.dev/leaderboards/benchmarks/reasoning-gpqa-diamond

Every model the catalog carries a published GPQA Diamond value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Reasoning |
| Measure | Multiple choice questions |
| Tasks | Expert-level science questions |
| Difficulty | Graduate-level scientific reasoning |
| Published by | [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://arxiv.org/abs/2311.12022) |

The hardest subset of GPQA featuring the most challenging graduate-level science questions. Sometimes reported separately from the standard GPQA benchmark.

## GPQA Diamond ranking

2 models with a published GPQA Diamond 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 | Multiple choice questions | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Sarvam 105B](https://modelchoice.dev/models/sarvam-105b) | Sarvam | 78.7 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Sarvam 30B](https://modelchoice.dev/models/sarvam-30b) | Sarvam | 66.5 | 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.
