Mbps and MB/s: differences and transfer time

Short answer

Mbps and MB/s measure different rates. Lowercase b means bits and uppercase B means bytes. Because one byte contains eight bits, convert Mbps to MB/s by dividing by 8:

MB/s = Mbps ÷ 8
Mbps = MB/s × 8

Therefore, 100 Mbps equals 12.5 MB/s, 600 Mbps equals 75 MB/s, and 1,000 Mbps equals 125 MB/s. These are exact unit conversions, not guaranteed download speeds.

Mbps versus MB/s at a glance

Capitalization matters: 1 MB/s is eight times 1 Mbps. Writing Mbit/s for megabits and MB/s for megabytes makes the distinction easier to see.

Quick Mbps to MB/s conversion table

To convert any other rate, use the bandwidth and transfer time calculator. It keeps the selected decimal or binary units visible and performs the calculation locally.

Why an actual download is usually slower

Dividing by eight gives the theoretical equivalence between units. It does not remove protocol headers, retransmissions, Wi-Fi interference, congestion, or limits imposed by the server, browser, CPU, or storage device.

A 1,000 Mbps link can carry 125 MB/s mathematically, but a real download may settle below that value. Use measured application throughput when estimating a real transfer. Do not subtract one universal overhead percentage: the loss depends on the network and workload.

Transfer time formula

Transfer time is data size divided by data rate. Both values must use matching units:

time in seconds = data in bits ÷ rate in bits per second

If the file size is in bytes and the rate is in bits per second, multiply the file size by eight first:

time in seconds = file size in bytes × 8 ÷ rate in bit/s

Alternatively, convert the rate to bytes per second and divide bytes by bytes per second.

Example: download 10 GB at 100 Mbps

Using decimal units, which are common for advertised network and storage capacities:

10 GB = 10,000,000,000 bytes
10,000,000,000 bytes × 8 = 80,000,000,000 bits
100 Mbps = 100,000,000 bit/s
80,000,000,000 ÷ 100,000,000 = 800 seconds

Eight hundred seconds is 13 minutes and 20 seconds. This is the ideal time at a constant 100 Mbps with no allowance for overhead or slowdown.

Ideal time for a 1 GB file

The following examples use exactly 1,000,000,000 bytes and a constant rate:

For an estimated efficiency of 85%, divide by 0.85. At 100 Mbps, the 10 GB example becomes approximately 941.18 seconds, or 15 minutes and 41 seconds. The 85% figure is an explicit assumption for the example, not a general rule.

MB and MiB are different too

Bits versus bytes and decimal versus binary prefixes are separate questions:

1 MB  = 1,000,000 bytes
1 MiB = 1,048,576 bytes
1 GB  = 1,000,000,000 bytes
1 GiB = 1,073,741,824 bytes

This means 600 Mbps is exactly 75 MB/s but approximately 71.5256 MiB/s. Dividing by eight resolves bits versus bytes; you must still confirm whether the prefix is decimal (MB, GB) or binary (MiB, GiB).

NIST documents binary prefixes to keep these quantities unambiguous. The guide to GB versus GiB covers static capacities and common drive sizes.

Throughput, goodput, and link speed

The negotiated link speed is not always the rate of useful application data. Throughput is a rate observed at a point in the system. Goodput focuses on useful application data and excludes items such as retransmissions or headers, depending on the measurement definition.

RFC 5166 distinguishes these network capacity metrics. TCP performance can also depend on round-trip time, window sizes, and loss, as discussed in RFC 6349. A slow disk or server can be the bottleneck even when the network has spare capacity.

A reliable four-step method

1. Identify whether the data size is in bits or bytes.
2. Identify whether the prefix is decimal (MB, GB) or binary (MiB, GiB).
3. Convert the rate and size to matching units.
4. Divide size by rate, then express the result in seconds, minutes, or hours.

Keep the units beside every value. If the final division is bit ÷ (bit/s), the result must be seconds.

Common mistakes

• Reading 600 Mbps as 600 MB/s and overestimating speed by a factor of eight.
• Treating MB/s and MiB/s as identical.
• Using the negotiated link rate as if it were sustained application goodput.
• Applying a fixed loss percentage without stating where it came from.
• Mixing decimal file sizes with binary rates during the same calculation.
• Converting the time to minutes while leaving the rate expressed per second.

What this calculation cannot tell you

A transfer-time estimate does not test a connection, diagnose Wi-Fi, predict a server’s load, or automatically account for latency, overhead, compression, and retransmissions. It relates size, rate, and time under assumptions that must be stated explicitly.

The Tools in a Tab calculator does the arithmetic in your browser. It does not download a test file or generate network traffic, so the values you enter are either a hypothesis or measurements obtained elsewhere.