Download data3/show_pricing.py from SciCode/dataset-builder: direct link, hf CLI and curl.
- Browser
- Download file 3.83 kB
-
https://huggingface.co/SciCode/dataset-builder/resolve/main/data3/show_pricing.py
- Command line
-
hf download hf://SciCode/dataset-builder/data3/show_pricing.py
-
curl -L -o show_pricing.py https://huggingface.co/SciCode/dataset-builder/resolve/main/data3/show_pricing.py
3.83 kB
| #!/usr/bin/env python3 | |
| """ | |
| Display OpenAI model pricing and cost estimates | |
| """ | |
| PRICING = { | |
| # GPT-5 series | |
| ("gpt-5.2", "GPT-5.2", "更高质量"): (1.75, 14.00), | |
| ("gpt-5.1", "GPT-5.1", "高质量"): (1.25, 10.00), | |
| ("gpt-5", "GPT-5", "高质量"): (1.25, 10.00), | |
| ("gpt-5-mini", "GPT-5 Mini", "较便宜"): (0.25, 2.00), | |
| ("gpt-5-nano", "GPT-5 Nano", "最便宜"): (0.05, 0.40), | |
| # GPT-5 Pro | |
| ("gpt-5.2-pro", "GPT-5.2 Pro", "顶级质量"): (21.00, 168.00), | |
| ("gpt-5-pro", "GPT-5 Pro", "顶级质量"): (15.00, 120.00), | |
| # GPT-4.1 series | |
| ("gpt-4.1", "GPT-4.1", "平衡选择"): (2.00, 8.00), | |
| ("gpt-4.1-mini", "GPT-4.1 Mini", "经济实惠"): (0.40, 1.60), | |
| ("gpt-4.1-nano", "GPT-4.1 Nano", "超低成本"): (0.10, 0.40), | |
| # GPT-4o series (recommended) | |
| ("gpt-4o", "GPT-4o ⭐", "质量优秀"): (2.50, 10.00), | |
| ("gpt-4o-mini", "GPT-4o Mini ⭐⭐", "性价比最高"): (0.15, 0.60), | |
| } | |
| def calculate_cost(input_price, output_price, input_tokens=2000, output_tokens=1500): | |
| """Calculate cost for given token counts""" | |
| input_cost = (input_tokens / 1_000_000) * input_price | |
| output_cost = (output_tokens / 1_000_000) * output_price | |
| return input_cost + output_cost | |
| def main(): | |
| print("=" * 90) | |
| print("OPENAI MODEL PRICING & COST ESTIMATES (Dec 2024)") | |
| print("=" * 90) | |
| print() | |
| # Assumptions | |
| input_tokens = 2000 | |
| output_tokens = 1500 | |
| print(f"Assumptions: {input_tokens:,} input tokens + {output_tokens:,} output tokens per problem") | |
| print() | |
| # Header | |
| print(f"{'Model':<25} {'Input':<12} {'Output':<12} {'Per Prob':<12} {'100 Probs':<12} {'1000 Probs':<12}") | |
| print("-" * 90) | |
| # Sort by cost per problem | |
| model_costs = [] | |
| for (model_id, model_name, desc), (input_price, output_price) in PRICING.items(): | |
| cost_per_problem = calculate_cost(input_price, output_price, input_tokens, output_tokens) | |
| model_costs.append((model_name, input_price, output_price, cost_per_problem, desc)) | |
| model_costs.sort(key=lambda x: x[3]) # Sort by cost per problem | |
| for model_name, input_price, output_price, cost_per_problem, desc in model_costs: | |
| cost_100 = cost_per_problem * 100 | |
| cost_1000 = cost_per_problem * 1000 | |
| print(f"{model_name:<25} ${input_price:<11.2f} ${output_price:<11.2f} " | |
| f"${cost_per_problem:<11.6f} ${cost_100:<11.2f} ${cost_1000:<11.2f}") | |
| print("-" * 90) | |
| print() | |
| # Recommendations | |
| print("RECOMMENDATIONS:") | |
| print(" ⭐⭐ gpt-4o-mini: Best cost-effectiveness for large batches") | |
| print(" ⭐ gpt-4o: Best quality-to-cost ratio for high-quality datasets") | |
| print(" 💰 gpt-4.1-nano or gpt-5-nano: Cheapest options for experiments") | |
| print(" 🏆 gpt-5-pro/gpt-5.2-pro: Premium options for critical applications") | |
| print() | |
| # Sample budgets | |
| print("SAMPLE BUDGETS:") | |
| print() | |
| recommended = [ | |
| ("gpt-5-nano", "Budget Testing", 0.05, 0.40), | |
| ("gpt-4.1-nano", "Low Budget", 0.10, 0.40), | |
| ("gpt-4o-mini", "Recommended ⭐", 0.15, 0.60), | |
| ("gpt-4.1-mini", "Mid Budget", 0.40, 1.60), | |
| ("gpt-4o", "High Quality ⭐", 2.50, 10.00), | |
| ] | |
| budget_amounts = [1.0, 5.0, 10.0, 20.0, 50.0] | |
| for model_id, label, input_p, output_p in recommended: | |
| cost_per = calculate_cost(input_p, output_p, input_tokens, output_tokens) | |
| print(f"\n{label} ({model_id}):") | |
| print(f" Cost per problem: ${cost_per:.6f}") | |
| counts = [int(budget / cost_per) for budget in budget_amounts] | |
| budget_str = " | ".join([f"${b}→{c:,}" for b, c in zip(budget_amounts, counts)]) | |
| print(f" Problems: {budget_str}") | |
| print() | |
| print("=" * 90) | |
| if __name__ == "__main__": | |
| main() | |