CodingDeprecated
Trust Score:
10

Mercury

Python code-synthesis benchmark that rewards both functional correctness and runtime efficiency against real solution distributions.

Launched: Refresh: static
Status Assessment (deprecated):

Mercury is a static 256-task public LeetCode-derived evaluation with no official maintained leaderboard or refreshed test set. Beyond remains useful for efficiency research, but the 2024 launch snapshot is no longer a frontier comparison.

Performance Timeline

Longitudinal progression of model scores against human baselines.

Performance & Historical Trajectory

Empirical score progression across model release dates and evaluation rounds.

Independent Vendor-Reported Human Baseline (70.0%)
ModelScoreDateSource TypeProvenance
CodeQwen1.5-7B base (Mercury-eval Overall Beyond, 5 samples)47.78%2024-04-18independentSource ↗
StarCoder2-15B base (Mercury-eval Overall Beyond, 5 samples)49.17%2024-02-28independentSource ↗
DeepSeek-Coder-33B base (Mercury-eval Overall Beyond, 5 samples)48.53%2023-11-02independentSource ↗
CodeLlama-34B base (Mercury-eval Overall Beyond, 5 samples)42.4%2023-08-24independentSource ↗

Human Baseline & Difficulty Horizon

Calibrated human reference points, specialist benchmarks, and ceiling thresholds.
Measured Human Score70.0%Domain Expert Baseline
Baseline Protocol & Interpretation

Human distribution baseline across Beyond@1 computational efficiency and correctness benchmarks.

Metric & Scoring Methodology

Verification protocols, aggregation formulas, and specialized metric variants.
Primary Metric:Beyond with 5 sampled solutions per task (%)
Scoring Engine:composite

Dataset & Compute Cost

Evaluation volume, public availability, API pricing, and local hardware requirements.
Total Dataset Size256Annotated evaluation items
Public Test SetPublicOpenly mirrored on repositories
Access GatingOpen AccessUnrestricted download
Evaluation LicenseOpen AccessDataset usage and redistribution terms
Frontier API Compute Cost:

$5 – $20 USD for full benchmark evaluation run on frontier APIs.

Recommended Local GPU Setup:

1x NVIDIA RTX 4090 (24GB) or A100 (40GB/80GB) via vLLM / SGLang

Official Dataset & Benchmark Files:Download / View Dataset Repository ↗

How to Run & Reproduce

Standardized evaluation protocols, CLI commands, and reproducible runner templates.
Prompt Regimezero-shot
Reasoning Modedirect
Sampling Temp0
Pass@k Budgetk = 1
Tools & SandboxPure Text
Scoring Verifiercomposite
Option AEleutherAI LM-Evaluation-Harness (Open-Weight Models)
lm_eval --model hf --model_args pretrained=<model_path> --tasks mercury --batch_size auto
Option BOpenCompass Evaluation Framework
opencompass --datasets mercury --models <model_config>
Python APIDeterministic Inference Loop Snippet
# Standard API Evaluation Loop
from openai import OpenAI

client = OpenAI()
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": prompt}],
    temperature=0.0,
)
Standardized Reporting Requirement:

When publishing results for Mercury, always report the exact prompt template, few-shot exemplar ordering, sampling temperature (temperature=0), maximum reasoning budget tokens, and the precise timestamped model snapshot ID.

Contamination & Memorization Analysis

Audit of pretraining exposure risks, memorization vectors, and refresh policies.
Overall Contamination Risk:HIGH
Refresh Cadence:

Static fixed snapshot

Test Set Exposure:

Public on web / HuggingFace