Datasetpublic

DropTheHQ/benchgecko-ai-pricing

BenchGecko AI Model Pricing 2026 Current pricing data for 413 AI models across 57 providers, with benchmark performance scores for cost-efficiency analysis. Sourced from BenchGecko. Why This Dataset Choosing an AI model is a price-performance tradeoff. This dataset enables: Cost-efficiency rankings (score per dollar) Provider pricing comparisons Open source vs proprietary cost analysis Budget optimization for production deployments Dataset Summary… See the full description on the dataset page: https://huggingface.co/datasets/DropTheHQ/benchgecko-ai-pricing.

sourceHugging Facecc-by-4.0updated 6mo agoView on Hugging Face

0likes19downloads

Dataset Card

BenchGecko AI Model Pricing 2026

Current pricing data for 413 AI models across 57 providers, with benchmark performance scores for cost-efficiency analysis. Sourced from BenchGecko.

Why This Dataset

Choosing an AI model is a price-performance tradeoff. This dataset enables:

—Cost-efficiency rankings (score per dollar)
—Provider pricing comparisons
—Open source vs proprietary cost analysis
—Budget optimization for production deployments

Dataset Summary

Metric	Value
Total models	413
Models with pricing	344
Open source models	235
Providers	57
Price range (input)	$0.01 - $150 per Mtok

Columns

Column	Type	Description
`model_slug`	string	Unique identifier
`model_name`	string	Display name
`provider`	string	Company (OpenAI, Anthropic, Google, etc.)
`provider_slug`	string	Provider identifier
`input_price_per_mtok`	float	Input cost in USD per million tokens
`output_price_per_mtok`	float	Output cost in USD per million tokens
`context_window`	int	Maximum context length in tokens
`is_open_source`	bool	Whether weights are publicly available
`release_date`	date	Release date
`model_type`	string	text, multimodal, etc.
`avg_benchmark_score`	float	Average across all benchmarks (0-100)
`num_benchmarks_tested`	int	How many benchmarks this model was tested on
`price_per_score_point`	float	Input price / avg score (lower = more efficient)

Usage

python

import pandas as pd

df = pd.read_csv("hf://datasets/DropTheHQ/benchgecko-ai-pricing/pricing.csv")

# Most cost-efficient models
efficient = df[df["input_price_per_mtok"].notna() & (df["input_price_per_mtok"] > 0)]
efficient = efficient.sort_values("price_per_score_point")
print(efficient[["model_name", "provider", "avg_benchmark_score", "input_price_per_mtok", "price_per_score_point"]].head(15))

# Cheapest models with score > 70
good_cheap = df[(df["avg_benchmark_score"] > 70) & (df["input_price_per_mtok"] > 0)]
print(good_cheap.nsmallest(10, "input_price_per_mtok")[["model_name", "provider", "avg_benchmark_score", "input_price_per_mtok"]])

# Average pricing by provider
by_provider = df[df["input_price_per_mtok"].notna()].groupby("provider").agg(
    models=("model_name", "count"),
    avg_input_price=("input_price_per_mtok", "mean"),
    avg_score=("avg_benchmark_score", "mean")
).round(2).sort_values("avg_input_price")
print(by_provider.head(15))

Source

Data from BenchGecko -- an independent AI tracking platform. Explore the full pricing calculator at benchgecko.ai/pricing.

License

CC BY 4.0

Citation

bibtex

@dataset{benchgecko_pricing_2026,
  title={BenchGecko AI Model Pricing 2026},
  author={BenchGecko},
  year={2026},
  url={https://benchgecko.ai},
  publisher={Hugging Face},
  license={CC BY 4.0}
}