# CVDP — Mercor evaluation (CVDP (Mercor))

> Hardware design, verification, and comprehension tasks. This table shows Mercor-run configurations for reference and is excluded from model rankings.

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

- Category: [Coding](/coding)
- Last updated: October 15, 2026 capture

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

We captured Mercor's CVDP results on October 15, 2026. The table preserves 25 published configurations, their source identifiers, reasoning settings, and disclosed pass@1 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 source page's selected Pass@1 metric. The task set, harness, tools, and grader belong to this Mercor run and may differ from another evaluator's results.

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 code generation: programmatic — a cocotb testbench runs against icarus verilog in docker and the task passes iff pytest passes. comprehension: bleu-2 overlap against the reference answer (correspondence) and an llm rating (q&a). and a judge of None for code generation; BLEU/ROUGE similarity and an LLM judge for comprehension. Its run-count label is k = 5. The linked result page takes precedence where its setup differs, as noted above.

- [Mercor CVDP results](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/)
- [Mercor APEX benchmark index](https://www.mercor.com/apex/)
- [Mercor methodology](https://www.mercor.com/apex/methodology/)
- [Source paper](https://arxiv.org/abs/2506.14074)
- [Upstream evaluation code](https://github.com/NVlabs/cvdp_benchmark)
- [Upstream dataset](https://huggingface.co/datasets/nvidia/cvdp-benchmark-dataset)

## About CVDP (Mercor)

- Tasks: 517 source-reported tasks
- Format: Pass@1
- Difficulty: Source-specific evaluation
- Paper: [CVDP evaluation results](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/)

The selected headline is Pass@1. 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.

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

## Leaderboard (25 configurations)

| Rank | Model | Configuration | Creator | Score |
|------|-------|---------------|---------|-------|
| 1 | [Opus 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,575 reported samples | Anthropic | 64.10% |
| 2 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | Anthropic | 62.60% |
| 3 | [Opus 5](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | Anthropic | 62.20% |
| 4 | [Fable 5.1](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · high reasoning · Published margin ±3.9 · 2,585 reported samples | Anthropic | 61.20% |
| 5 | [Sonnet 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,568 reported samples | Anthropic | 60.60% |
| 6 | [Grok 4.6](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · xhigh reasoning · Published margin ±3.9 · 2,585 reported samples | xAI | 58.00% |
| 7 | [Haiku 5.5](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,576 reported samples | Anthropic | 57.00% |
| 8 | [Sonnet 5](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,541 reported samples | Anthropic | 56.40% |
| 9 | [Kimi K3](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | Kimi | 55.50% |
| 10 | [GPT-5.6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,585 reported samples | OpenAI | 55.40% |
| 11 | [Qwen3.8-Max](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · xhigh reasoning · Published margin ±4.1 · 2,518 reported samples | Alibaba | 55.20% |
| 12 | [GPT-6.1 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4.2 · 2,567 reported samples | OpenAI | 55.10% |
| 13 | [GPT-6 Astra](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4.2 · 2,585 reported samples | OpenAI | 54.90% |
| 14 | [Grok 4.7](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · xhigh reasoning · Published margin ±3.9 · 2,549 reported samples | xAI | 54.90% |
| 15 | [Muse Spark 1.3](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · xhigh reasoning · Published margin ±3.9 · 2,585 reported samples | Meta | 54.30% |
| 16 | [GPT-5.6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · xhigh reasoning · Published margin ±4.1 · 2,585 reported samples | OpenAI | 54.10% |
| 17 | [DeepSeek-V4-Pro-0813](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | DeepSeek | 53.10% |
| 18 | [GLM-5.3](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | Zhipu | 53.10% |
| 19 | [GPT-6 Sol](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,585 reported samples | OpenAI | 52.70% |
| 20 | [Gemini 3.1 Pro](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · high reasoning · Published margin ±4 · 2,585 reported samples | Google | 51.30% |
| 21 | [GPT-5.6 Terra](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.8 · 2,585 reported samples | OpenAI | 50.90% |
| 22 | [GPT-6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±4 · 2,585 reported samples | OpenAI | 50.80% |
| 23 | [Gemini 3.8 Flash](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · high reasoning · Published margin ±3.8 · 2,585 reported samples | Google | 50.70% |
| 24 | [GPT-5.6 Luna](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · max reasoning · Published margin ±3.9 · 2,585 reported samples | OpenAI | 48.40% |
| 25 | [MiniMax-M3](https://www.mercor.com/apex/oss-benchmarks/oss-cvdp-leaderboard/) | Terminus via Harbor (1 hour time limit) for code generation; single-shot for comprehension · high reasoning · Published margin ±3.7 · 2,585 reported samples | MiniMax | 44.80% |

## FAQ

### What does CVDP (Mercor) measure?

Hardware design, verification, and comprehension tasks. This table shows Mercor-run configurations for reference and is excluded from model rankings.

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

Opus 5.5 currently leads the published CVDP (Mercor) snapshot with a score of 64.10%.

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

The October 15, 2026 capture contains 25 source configurations.

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

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