# Git Repo Size & LFS Cost Calculator Price Git LFS storage and bandwidth against your plan allowance and project the bill as the repo grows. --- - **Canonical URL:** https://dothecalculation.com/calculators/git-repo-size-cost-calculator - **Category:** AI & Tech Development - **Publisher:** Do The Calculation (https://dothecalculation.com) - **Cost:** Free, no account or sign-up required - **Privacy:** Runs entirely in the browser; inputs are never sent to a server - **Methodology:** https://dothecalculation.com/methodology --- ## What Does Your Git LFS Repository Actually Cost? Price LFS storage and bandwidth against your plan allowance, project the bill as the repository grows, and find out how few clones it takes to burn a month of bandwidth. - Storage and bandwidth priced separately, because only one of them is usually the problem - Month-by-month projection as LFS objects accumulate - The clones-to-exhaust figure that explains most surprise bills ## Quick Answer — Why Is Git LFS Billing So Surprising? Because the expensive part is invisible. Storage is a number you can see in the repository; bandwidth is generated every time anyone clones, and every time a CI job checks out without a cache. On most teams, bandwidth is the overwhelming majority of the bill. GitHub bills Git LFS on metered usage: an included allowance per plan, then a per-GiB rate above it. **Free and Pro accounts include 10 GiB of storage and 10 GiB of bandwidth; Team and Enterprise Cloud include 250 GiB of each.** Pre-paid data packs have been retired in favour of metered billing. **A mid-size repository: 40 GB of LFS objects, growing 3 GB a month, cloned 220 times a month at 2.5 GB per clone, on a 10 GiB allowance:** • Billable storage — **30 GB**, costing **$2.10 a month** • Billable bandwidth — **540 GB**, costing **$47.25 a month** • Total now — **$49.35 a month**, of which bandwidth is **96 percent** • Twelve months later — **$51.66 a month**, **$606.06** in total across the year • Clones needed to exhaust the allowance — **4.0** Four clones. That is the whole story: a repository with a 2.5 GB working set burns its entire monthly bandwidth allowance before five people have checked it out. ## How to Use This Calculator: A 40 GB Asset Repository Enter the two size numbers separately. **LFS storage** is what the large-file store holds — here 40 GB. **Plain Git history** is the ordinary repository, here 1.2 GB. Git history compresses well and is almost never the cost driver, but keeping it visible stops you from blaming the wrong thing. Set the growth rate. Three gigabytes a month of new or revised assets is unremarkable for a game project or a design repository, and after twelve months it takes the store to **73 GB** — nearly double, with no decision having been made along the way. Now the traffic. Twenty developers cloning fresh once a month, plus 200 CI runs that check out LFS objects, is 220 checkouts. At **2.5 GB** of LFS data each that is **550 GB** of bandwidth a month, of which **540 GB** is billable. At $0.07 per GB for storage and $0.0875 per GB for bandwidth, the monthly bill is **$49.35** — $2.10 of storage and $47.25 of bandwidth. Twelve months of growth takes the monthly figure to **$51.66** and the cumulative total to **$606.06**. Change one input to see where the leverage is. Cutting clone traffic in half saves about $24.06 a month. Halving the stored data saves about $1.40. The two levers are not remotely comparable. ## The Formulas This Calculator Uses **Billable storage** = max(0, LFS storage − included allowance). **Monthly bandwidth** = clones and CI checkouts × GB of LFS data per checkout. **Billable bandwidth** = max(0, monthly bandwidth − included allowance). **Monthly cost** = billable storage × storage rate + billable bandwidth × bandwidth rate. **Storage in month n** = current storage + monthly growth × n, with the bill recomputed each month. **Clones to exhaust the allowance** = included allowance ÷ GB per clone. This is the figure worth writing on a whiteboard. The rates are editable because they are the volatile part. GitHub's documentation confirms the plan allowances and that billing is metered per GiB, and the defaults here are the commonly quoted per-GiB figures — but check them against your own billing page before you rely on a total. The allowances and the arithmetic are stable; the price per gigabyte is not. ## A Second Example: A Small Repository and a Large One **The small case.** An 8 GB LFS store growing 0.5 GB a month, cloned 30 times at 0.9 GB each, on the 10 GiB Free and Pro allowance. Storage is entirely inside the allowance and costs nothing. Bandwidth is 27 GB, of which 17 GB is billable, costing **$1.49 a month**. Twelve months later the store has reached 14 GB and the monthly bill is **$1.73** — trivial, but no longer zero, and it crossed that line in month five without anyone noticing. **The large case.** A 600 GB LFS store growing 25 GB a month, cloned 900 times a month at 12 GB per checkout, on a Team plan with a 250 GiB allowance. Bandwidth is **10,800 GB** a month, of which 10,550 GB is billable. The monthly bill is **$947.62** — a run rate of **$11,854.50 a year**, and 97 percent of it is checkout traffic. The large case is where the engineering fix pays for itself many times over. Caching LFS objects on CI runners, using partial or shallow clones, and setting the smudge filter to skip files a job never touches can remove most of that 10,800 GB. None of it changes the stored data by a single byte. It is also the case where the question should be asked properly: 12 GB per checkout means the repository is carrying an asset library, not source code. Object storage with pointers in the repository often costs an order of magnitude less. ## How to Actually Reduce the Bill **Cache LFS objects in CI.** This is the single highest-leverage change on almost every repository, because CI is usually the majority of checkout traffic and it re-downloads the same objects every run. A cache keyed on the LFS object ids turns hundreds of full pulls a month into a handful. **Skip the smudge filter on jobs that do not need the files.** A lint job, a unit test suite, or a documentation build usually never touches the binary assets, and setting the environment variable that skips LFS smudging makes those checkouts free. **Prune what nobody needs.** Every version of every binary stays in LFS forever unless someone removes it. A repository that has been running for three years is usually carrying several generations of assets that no branch references. **Question whether LFS is the right home at all.** LFS earns its cost for files that are genuinely versioned alongside code and need to arrive with a checkout. For a large asset library that changes independently, object storage with a manifest in the repository is usually far cheaper — the [cloud storage plan comparison calculator](/calculators/cloud-storage-plan-comparison-calculator) prices that alternative on cost per TB, and the [data storage calculator](/calculators/data-storage-calculator) sizes the library before you decide. **Watch the second bill.** LFS and CI minutes are billed separately and driven by the same runs, so a pipeline change that halves checkouts often halves both. The [CI/CD minutes cost calculator](/calculators/cicd-minutes-cost-calculator) covers the other side of that ledger. ## Limitations The per-GiB rates are inputs, not verified constants. GitHub's documentation confirms the plan allowances and that billing is metered, but does not publish the per-GiB figures on the same page, so the defaults here come from secondary sources. Check your billing page and replace them before quoting a number to anyone. Storage is billed on an hourly usage rate in practice, and this model treats it as a monthly snapshot. For a repository whose size changes gradually that is close enough; for one that spikes and shrinks it will be off. The model assumes every checkout pulls the full LFS working set. Real traffic is messier: shallow clones, sparse checkouts, cached runners, and developers who already have most objects locally all pull less. Enter an average GB per checkout that reflects what actually happens rather than the size of the store. GitHub is not the only forge, and every host prices LFS differently — some bundle it, some meter it, some cap it. The arithmetic here applies anywhere; the allowances and rates do not. ## Related Calculators Repository storage is one line in a larger infrastructure bill. The [data storage calculator](/calculators/data-storage-calculator) sizes a file library in GB and TB before you decide where to put it, and the [cloud storage plan comparison calculator](/calculators/cloud-storage-plan-comparison-calculator) compares plans on cost per TB and cost per GB actually stored — the comparison that decides whether LFS or object storage is the right home for an asset directory. The [CI/CD minutes cost calculator](/calculators/cicd-minutes-cost-calculator) prices the pipeline runs that generate most of the checkout traffic, and the [compression ratio calculator](/calculators/compression-ratio-calculator) shows what compressing assets before they enter the store would buy. ## Frequently asked questions ### How much does Git LFS cost? GitHub includes 10 GiB of LFS storage and 10 GiB of bandwidth on Free and Pro plans, and 250 GiB of each on Team and Enterprise Cloud, then bills per GiB above that. A 40 GB repository cloned 220 times a month at 2.5 GB per clone costs about $49.35 a month at the default rates, of which 96 percent is bandwidth. ### Why is my LFS bandwidth bill so much larger than my storage bill? Because storage is charged once and bandwidth is charged every time anyone pulls. A 40 GB store costs $2.10 a month above a 10 GiB allowance; the 550 GB of monthly clone and CI traffic against the same repository costs $47.25. Bandwidth is over twenty times the storage cost in that example. ### How many clones use up my LFS bandwidth allowance? Divide the allowance by the LFS data per checkout. On a 10 GiB allowance with a 2.5 GB working set, four full clones exhaust the month. On a Team plan with 250 GiB and a 12 GB working set, it is about twenty-one. CI runs count, which is why the number is usually reached in the first week. ### Are GitHub LFS data packs still available? No. GitHub's billing documentation states that pre-paid data packs have been removed and replaced with metered billing, so usage above the plan allowance is charged per GiB rather than bought in blocks of 50 GB. ### How do I reduce Git LFS costs? Attack bandwidth first, because it is almost always the bill. Cache LFS objects on CI runners, use shallow or partial clones, and skip the smudge filter on jobs that never touch the binaries. Pruning stored objects helps too, but on the example repository halving the storage saves $1.40 a month against $24.06 for halving the traffic. ### Should I use Git LFS or object storage? LFS earns its cost for binaries that are genuinely versioned alongside the code and need to arrive with a checkout. For a large asset library that changes on its own schedule, object storage with a manifest in the repository is usually far cheaper. If your checkouts pull 12 GB each, you are carrying an asset library rather than source code. ### Does plain Git history count toward LFS billing? No. Ordinary Git objects live in the repository and are covered by the plan, and they compress well enough that even an old project is usually well under a gigabyte. Only files tracked by LFS count against the LFS storage and bandwidth allowances, which is why this calculator asks for the two figures separately. ### How fast will my repository grow? Faster than expected, because nothing ever leaves. Every revision of every binary stays in LFS unless it is actively pruned, so a repository adding 3 GB a month is 36 GB larger a year later whether or not anyone still needs those files. Project the horizon you actually plan over rather than the current month. ## Related concepts - **LFS Bandwidth** — Data downloaded from the large-file store on every clone and CI checkout. Invisible in the repository and usually the overwhelming majority of the bill. - **Plan Allowance** — The included storage and bandwidth before metering starts: 10 GiB of each on Free and Pro, 250 GiB of each on Team and Enterprise Cloud. - **Clones to Exhaust** — The allowance divided by the LFS data per checkout. On a 10 GiB allowance and a 2.5 GB working set, four checkouts use the whole month. ## Related guides - [How Much Cloud Storage Do You Actually Need?](https://dothecalculation.com/blog/tech/cloud-storage-planning-guide) — Why video dominates storage far more than photo count suggests, and how to compare Google One, iCloud+, and Dropbox pricing honestly using official 2026 rates. - [How to Use Do The Calculation Calculators: A Practical Step-by-Step Guide](https://dothecalculation.com/blog/site-guides/how-to-use-calculators) — Learn the fastest reliable workflow for using Do The Calculation calculators, reading results, checking formulas, and using save, print, share, and export actions correctly. ## Related calculators - [CI/CD Minutes Cost Calculator](https://dothecalculation.com/calculators/cicd-minutes-cost-calculator) — Price GitHub Actions minutes with per-job rounding, OS multipliers, and the included allowance on your plan. - [Website Hosting Cost Comparison Calculator](https://dothecalculation.com/calculators/website-hosting-cost-calculator) — Break a hosting bill into base fee, bandwidth overage, and storage overage to see what's driving your monthly cost. - [Cache Hit Rate & CDN Cost Savings Calculator](https://dothecalculation.com/calculators/cache-hit-rate-calculator) — Project origin server load reduction, bandwidth savings, and network cost return on investment when using a CDN, instantly and free. - [Cloud Storage & Egress Cost Calculator](https://dothecalculation.com/calculators/cloud-egress-cost-calculator) — Compare bandwidth transfer and data egress costs across AWS, Google Cloud, Azure, and Cloudflare R2 storage providers instantly. - [Cloud Storage Plan Cost Comparison Calculator](https://dothecalculation.com/calculators/cloud-storage-plan-comparison-calculator) — Compare cloud storage plans on cost per TB, cost per GB you actually store, and five-year total. - [Database Storage Growth Projection Calculator](https://dothecalculation.com/calculators/database-storage-growth-calculator) — Project database size from rows and bytes, then convert it into the disk you actually have to provision. --- _Plan allowances and the fact that Git LFS is billed on metered usage rather than pre-paid data packs come from GitHub's own billing documentation. The per-GiB rates are inputs with commonly quoted defaults rather than verified constants, because GitHub does not publish them alongside the allowances — check your billing page before relying on a total. Storage is treated as a monthly snapshot where the real billing uses an hourly usage rate, and every checkout is assumed to pull the full LFS working set, so enter an average that reflects your actual clone and CI behaviour._ --- _Source: [Do The Calculation](https://dothecalculation.com/calculators/git-repo-size-cost-calculator). Quote freely with attribution and a link to this page._