Concurrents → RPS & Gbps calculator

Plan scale (not Little’s Law)

This calculator uses Slammed.st plan capacity estimates, not latency-based Little’s Law:

RPS ≈ concurrents × 60,000

Gbps ≈ concurrents × 30

So 1 concurrent ≈ 60k RPS ≈ 30 Gbps of generator capacity class. That is what you buy on the plan — peak generation headroom toward a verified target. Achieved goodput still depends on method, path, CDN, and how the target responds under load.

For latency ↔ in-flight workers math, see the blog: Concurrent connections vs RPS.

Interactive calculator

Worked examples

• 1 concurrent: ≈ 60,000 RPS and ≈ 30 Gbps (Trial-class ceiling).
• 2 concurrents: ≈ 120,000 RPS and ≈ 60 Gbps (Basic-class).
• 4 concurrents: ≈ 240,000 RPS and ≈ 120 Gbps (Pro-class).
• Target 300,000 RPS: concurrents ≈ 300000 ÷ 60000 = 5.
• Target 150 Gbps: concurrents ≈ 150 ÷ 30 = 5.

Always report plan concurrents, achieved RPS/Gbps, latency, and errors on systems you own or are authorized to test.

Caveats

Plan scale is a generator estimate. Real runs need headroom for queues, retries, think time, and protection layers. High RPS goals can fail on the target long before the generator is saturated.

Use abort criteria, step ramps, and verified-target checks. Glossary: RPS, concurrent connections, throughput.

Use it in authorized tests

Estimate the concurrents you need, pick a matching plan on pricing, verify the target, then ramp on Slammed.st.

Start a free trial · Authorized load testing hub · Tools matrix.

Using the calculator without fooling yourself

The concurrent → RPS / Gbps helper estimates generator class on Slammed.st plans. It does not promise your origin will sustain that goodput. Treat the output as a shopping and ramp-planning aid: pick a plan with headroom, then let verified-target runs reveal the real plateau under abort rules.

Latency-linked Little’s Law thinking still matters for interpreting open workers vs completed work — see the blog concurrent vs RPS deep dive and glossary entries for RPS and concurrent connections.

Suggested planning sequence

1. Write the target RPS or Gbps class for the hypothesis (with scenario name).
2. Convert to concurrents here; confirm the matching tier on pricing.
3. Verify the target; smoke far below the estimate.
4. Ramp toward the estimate only while latency and errors stay inside abort criteria.
5. Report achieved throughput separately from plan generator class.

Policy: authorized load testing. Trial: register. Automation: docs.

When not to trust a back-of-envelope conversion

Do not use plan scale math as proof for executives without a Hub report. Do not size production peaks solely from Trial-class experiments. If your question is purely script VU math for CI, keep that model next to application latency — this calculator is oriented to Slammed.st capacity classes.

Index of terms: glossary. Tooling matrix: load testing tools matrix.

Related

Resources hub · Glossary · Tools matrix · RPS calculator · Useful network tools · Blog · Pricing · Start free