# GSM8K leaderboard

Mathematics 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/math-gsm8k

Grade School Math 8K. Every model the catalog carries a published GSM8K value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Mathematics |
| Measure | Exact match |
| Tasks | Grade-school math word problems |
| Difficulty | Grade-school math |
| Published by | [Training Verifiers to Solve Math Word Problems](https://arxiv.org/abs/2110.14168) |

GSM8K tests multi-step arithmetic in grade-school word problems. OpenAI introduced the dataset in 2021; a high score on this fixed test is one piece of evidence about mathematical reasoning.

## GSM8K ranking

7 models with a published GSM8K 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 | Exact match | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [OrcaCyber-Nano-1.0](https://modelchoice.dev/models/orcacyber-nano-1-0) | OrcaRouter | 97.5 | Observed 2026-11-21 · source benchlm:models |
| 2 | [Ternary Bonsai 2 27B](https://modelchoice.dev/models/ternary-bonsai-2-27b) | Prism ML | 96.7 | Observed 2026-11-21 · source benchlm:models |
| 3 | [Llama 3.1 70B](https://modelchoice.dev/models/llama-3-1-70b) | Meta | 95.1 | Observed 2026-11-21 · source benchlm:models |
| 4 | [Celeris-1](https://modelchoice.dev/models/celeris-1) | Celeris | 93.7 | Observed 2026-11-21 · source benchlm:models |
| 5 | [Phi-3.5-mini](https://modelchoice.dev/models/phi-3-5-mini) | Microsoft | 86.2 | Observed 2026-11-21 · source benchlm:models |
| 6 | [Soofi S 30B-A3B](https://modelchoice.dev/models/soofi-s-30b-a3b) | Soofi Project | 86.1 | Observed 2026-11-21 · source benchlm:models |
| 7 | [Llama 3.1 8B](https://modelchoice.dev/models/llama-3-1-8b) | Meta | 84.5 | 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.
