Multi-Cloud Vendor Price Comparison Script

Comparing Cloud Vendor Costs with Python

"Just how much will the cloud cost over a full year?" is the question that always comes up during mid-year budget reviews or when a new project kicks off. Checking each vendor's pricing page and tallying manually is slow and error-prone. This article provides an extensible Python script that stores per-vendor unit prices in a dictionary, computes monthly and annual costs in one go, and sorts the results automatically. The script doesn't aim for precision — it standardizes the comparison so the same CPU, RAM, and term sit on one table, and the expensive option stands out immediately.

The script

plans = {
    "aliyun":  {"cpu": 18, "ram": 7,   "base": 10},
    "tencent": {"cpu": 17, "ram": 6.5, "base": 9},
    "vultr":   {"cpu": 10, "ram": 4.5, "base": 4},
    "aws":     {"cpu": 15, "ram": 6,   "base": 8},
    "hetzner": {"cpu": 6,  "ram": 3,   "base": 3},
}

def monthly_cost(cpu_core: int, ram_gb: int) -> dict:
    result = {}
    for name, p in plans.items():
        result[name] = round(p["base"] + p["cpu"] * cpu_core + p["ram"] * ram_gb, 2)
    return dict(sorted(result.items(), key=lambda x: x[1]))

def yearly_cost(cpu_core: int, ram_gb: int) -> dict:
    monthly = monthly_cost(cpu_core, ram_gb)
    return {name: round(cost * 12, 2) for name, cost in monthly.items()}

print(monthly_cost(2, 4))
print(yearly_cost(2, 4))

Reference: official pricing pages of each vendor (always trust the live quotes before ordering)

Adding more vendors

Add a line to the plans dictionary: fill in the per-core price, per-GB-RAM price, and base fee for each vendor. For DigitalOcean, write "do": {"cpu": 10, "ram": 5, "base": 4}. If a vendor uses tiered pricing (first-year discounts, volume pricing), replace base with a function or add conditions inside monthly_cost. To add more vendors, just keep appending — the sorting and aggregation logic needs no changes.

Example output

For a 2-core / 4GB configuration, the script prints:

monthly: {'hetzner': 27.0, 'vultr': 42.0, 'aws': 62.0, 'tencent': 69.0, 'aliyun': 74.0}
yearly:  {'hetzner': 324.0, 'vultr': 504.0, 'aws': 744.0, 'tencent': 828.0, 'aliyun': 888.0}

The price parameter table

A real scenario

A team needs a three-year budget for a content site running on 2 cores / 4GB with 1TB of monthly traffic. The script's annual figures range from hetzner's 324 to aliyun's 888 — a spread of more than 1,600 over three years. That gap is large enough to drive the server selection decision, especially when cost-effectiveness is an explicit requirement. The scenario also shows that if most of the budget goes to traffic, folding each vendor's traffic-billing rules into the comparison is closer to reality than comparing spec prices alone.

Feeding the script into a monthly report

Budget comparison shouldn't be a one-off exercise. Pipe yearly_cost output into your monthly report and run it every month to watch how each vendor's cost moves with configuration changes. It's straightforward: have the script write CSV via Python's csv module, then hand it to a reporting tool or a simple table template. When a vendor adjusts prices, the report reflects it right away instead of you finding out at renewal time. Show both a three-month trend and a yearly estimate in the report so management can judge cost direction at a glance.

Annual budgets and reserved instances

Annual cost here is 12 full-price months. For stable workloads, reserved instances or yearly plans typically save another 30%+; write the discount factor into the script, e.g. multiply by 0.7 in yearly_cost, to get a number closer to the real contract price. When purchasing, weigh monthly flexibility against annual discounts together.

Keeping the price data fresh

1. Trust live quotes over anything else. The numbers in the script are approximations; plug in current vendor promotions and the actual configuration before ordering.
2. Fold in the hidden costs. Egress traffic, snapshot backups, and load balancers are all extra — the breakdown in the server cost calculator covers this.
3. Re-check regularly. Cloud pricing changes often; update the parameter table quarterly so long-term decisions don't rest on stale numbers.
4. Mind regional differences. The same vendor can charge 20-40% more in one region than another — put the target region into the comparison too.

FAQ

Can I place an order based on this output? No. It assumes linear pricing, while real plans have tiers, promos, and regional differences; always check the live official quotes before ordering. Why do Alibaba and Tencent look more expensive than overseas vendors? These are simplified parameters, not real quotes; domestic vendors often have new-user discounts and annual plans, so fold promotions into the comparison. Is annual cost the same as one-time yearly payment? No — this sums 12 full-price months; yearly payment usually has a discount, so multiply in the discount factor. Can the output render on a web page? Yes — serialize the results to JSON and render them with a table component, or generate a static HTML page on a cron schedule. Can the USD figures be converted to CNY? Yes — multiply by an exchange-rate factor before output, or rewrite the prices in plans to CNY quotes; just keep the basis consistent.

Notes

• The script assumes a linear per-core/per-GB model; real plans often use tiered pricing, so check the bundle price directly for large machines.
• Annual cost here is 12 full-price months, ignoring reserved-instance discounts. For stable workloads, reserved instances typically save another 30%+.
• This is a comparison tool, not an ordering tool. The final decision should follow the vendor contract and the server selection guide.

For more head-to-head comparisons, browse the server selection category.