# GDPval — Mercor evaluation (GDPval (Mercor))

> Producing workplace deliverables from reference files. This table shows Mercor-run configurations for reference and is excluded from model rankings.

Canonical page: https://benchlm.ai/benchmarks/mercorgdpval

- Category: [Agentic](/agentic)
- Last updated: October 15, 2026 capture

Evaluation results published by [Mercor](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/). Published evaluation aggregates only; task contents and dataset license grants are not included.

We captured Mercor's GDPval results on October 15, 2026. The table preserves 45 published configurations, their source identifiers, reasoning settings, and disclosed mean score values. These are results from Mercor's evaluation setup; we did not rerun them.

This dated source table is display only and does not enter overall or category rankings. Source configurations retain their external identity, including Pro descriptors and different effort settings.

The headline is the published mean score, which can include partial credit. It must not be read as a strict task-completion rate.

The row labels retain published error margins and sample counts when supplied. Missing margins and counts remain unreported; a sample count is not used to reconstruct a task denominator. Except where the dedicated source defines an interval, the confidence level is unreported. Equal displayed scores and overlapping margins do not establish statistical ties. Capture time does not establish individual evaluation dates.

The general methodology lists llm-as-a-judge (lmaaj) and a judge of Gemini 3 Flash Preview. Its run-count label is k = 4. The linked result page takes precedence where its setup differs, as noted above.

- [Mercor GDPval results](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/)
- [Mercor APEX benchmark index](https://www.mercor.com/apex/)
- [Mercor methodology](https://www.mercor.com/apex/methodology/)
- [Source paper](https://arxiv.org/abs/2510.04374)
- [Upstream dataset](https://huggingface.co/datasets/openai/gdpval)

## About GDPval (Mercor)

- Tasks: 213 source-reported tasks
- Format: Mean Score
- Difficulty: Source-specific evaluation
- Paper: [GDPval evaluation results](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/)

The selected headline is Mean Score. Source model identifiers, effort settings, published margins, and reported sample counts remain attached to the source configuration. Public-set results and Mercor task extensions occupy separate tables.

GDPval (Mercor) is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (45 configurations)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [Opus 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.5 · 832 reported samples | Anthropic | 88.30% |
| 2 | [DeepSeek-V4.1-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.5 · 852 reported samples | DeepSeek | 86.70% |
| 3 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.7 · 847 reported samples | OpenAI | 86.70% |
| 4 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.7 · 851 reported samples | Anthropic | 86.60% |
| 5 | [Sonnet 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.7 · 819 reported samples | Anthropic | 86.50% |
| 6 | [Grok 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±1.6 · 850 reported samples | xAI | 86.00% |
| 7 | [Opus 5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.9 · 854 reported samples | Anthropic | 85.60% |
| 8 | [Haiku 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.6 · 834 reported samples | Anthropic | 85.50% |
| 9 | [Gemini 3.7 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±1.7 · 849 reported samples | Google | 85.20% |
| 10 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±2.1 · 868 reported samples | OpenAI | 84.80% |
| 11 | [Opus 4.8](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.7 · 852 reported samples | Anthropic | 84.50% |
| 12 | [Gemini 3.6 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±1.7 · 852 reported samples | Google | 83.80% |
| 13 | [Muse Spark 1.2](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±2 · 845 reported samples | Meta | 83.80% |
| 14 | [GPT-6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±1.9 · 852 reported samples | OpenAI | 83.50% |
| 15 | [Gemini 3.8 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±1.6 · 848 reported samples | Google | 83.50% |
| 16 | [Qwen3.8-Max](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±2 · 852 reported samples | Alibaba | 83.30% |
| 17 | [Muse Spark 1.1](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±2 · 851 reported samples | Meta | 82.40% |
| 18 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±2.7 · 852 reported samples | Anthropic | 81.40% |
| 19 | [GPT-6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±2.2 · 852 reported samples | OpenAI | 80.10% |
| 20 | [GPT-5.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±5.2 · 852 reported samples | OpenAI | 79.90% |
| 21 | [Opus 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.1 · 852 reported samples | Anthropic | 79.80% |
| 22 | [Sonnet 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.2 · 852 reported samples | Anthropic | 79.00% |
| 23 | [Grok 4.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.2 · 852 reported samples | xAI | 78.80% |
| 24 | [Kimi K3](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.1 · 852 reported samples | Kimi | 78.80% |
| 25 | [Fable 5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.1 · 852 reported samples | Anthropic | 78.60% |
| 26 | [DeepSeek-V4-Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5 · 852 reported samples | DeepSeek | 78.50% |
| 27 | [GLM-5.2](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.2 · 852 reported samples | Zhipu | 78.20% |
| 28 | [Sonnet 5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.2 · 852 reported samples | Anthropic | 77.60% |
| 29 | [GPT-5.6 Terra](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.2 · 852 reported samples | OpenAI | 77.10% |
| 30 | [GPT-5.6 Sol (Pro)](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.3 · 852 reported samples | OpenAI | 75.60% |
| 31 | [MiniMax-M3](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.3 · 852 reported samples | MiniMax | 74.80% |
| 32 | [DeepSeek-V4-Pro](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.1 · 852 reported samples | DeepSeek | 74.30% |
| 33 | [GPT-5.6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · max reasoning · Published margin ±5.5 · 852 reported samples | OpenAI | 72.50% |
| 34 | [Gemini 3.5 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.2 · 852 reported samples | Google | 72.40% |
| 35 | [DeepSeek-V3.2](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · Published margin ±2.1 · 852 reported samples | DeepSeek | 71.90% |
| 36 | [Qwen3.5](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · Published margin ±2.2 · 851 reported samples | Alibaba | 71.00% |
| 37 | [Kimi K2.7 Code](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.4 · 852 reported samples | Kimi | 70.10% |
| 38 | [Gemini 3.1 Pro](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.5 · 852 reported samples | Google | 68.50% |
| 39 | [MiniMax-M2.7](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.3 · 852 reported samples | MiniMax | 68.30% |
| 40 | [Inkling](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.6 · 852 reported samples | Thinking Machines | 65.40% |
| 41 | [Nemotron 3 Ultra](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±3.1 · 852 reported samples | NVIDIA | 63.50% |
| 42 | [GPT-5.4](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · xhigh reasoning · Published margin ±5.7 · 852 reported samples | OpenAI | 60.00% |
| 43 | [Kimi K2 (Thinking)](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.6 · 852 reported samples | Kimi | 54.10% |
| 44 | [Gemma 4 31B](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · Published margin ±5.8 · 852 reported samples | Google | 52.70% |
| 45 | [GPT-OSS-120B](https://www.mercor.com/apex/oss-benchmarks/oss-gdpval-leaderboard/) | Stirrup · high reasoning · Published margin ±5.5 · 852 reported samples | OpenAI | 38.80% |

## FAQ

### What does GDPval (Mercor) measure?

Producing workplace deliverables from reference files. This table shows Mercor-run configurations for reference and is excluded from model rankings.

### Which model leads the published GDPval (Mercor) snapshot?

Opus 5.5 currently leads the published GDPval (Mercor) snapshot with a score of 88.30%.

### How many models are evaluated on GDPval (Mercor)?

The October 15, 2026 capture contains 45 source configurations.

### Does GDPval (Mercor) affect BenchLM's overall score?

Not directly. GDPval (Mercor) is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.