{'TRIAL_CSS'} Skip to main content
// NEXOLOAD TITAN · COMING FALL 2026 · 10,000+ VUs

10,000+ VUs
One Machine.
Zero Extra Infrastructure.

NexoLoad Titan replaces the OS thread engine with async coroutines — pushing the virtual user ceiling from ~200 to 10,000+ on a single machine. Same CLI as Pro. Same flags. Just more power. Targeting 2,824+ TPS via async coroutine architecture.

Request Pre-Sales Demo Compare All Editions
NexoLoad
10,000+ Concurrent Virtual Users From a single machine
2,824 Target TPS Async engine design target · Fall 2026
83ms Target P95 Projected async coroutine performance
10KB RAM per Virtual User vs ~1,500 KB for OS threads

Real baseline — NexoLoad Lite · AWS Lambda (us-east-1) · controlled benchmark:  20 VUs: 118.5 TPS · P95 182.6ms · P99 212.8ms · 0% errors  |  50 VUs: 268.6 TPS · P95 208.2ms · P99 320.0ms · 0% errors  · Titan targets >10× the 50 VU throughput via async architecture.

Under the Hood

asyncio + aiohttp.
The Engine That Changes Everything.

NexoLoad Pro uses one OS thread per virtual user. Threads work up to ~200 VUs — then the OS scheduler overhead and memory usage become the bottleneck, not your API. Titan swaps threads for async coroutines. A coroutine uses ~10 KB versus ~1,500 KB for a thread. The event loop runs thousands concurrently on a single core.

Metric Pro (threads) Titan (async)
Max VUs~20010,000+
RAM per VU~1,500 KB~10 KB
RAM @ 1,000 VUs~1.5 GB~10 MB
Enginethreading.Threadasyncio + aiohttp
DependenciesZero (stdlib)aiohttp (bundled)
Zero-install✓✓ (aiohttp compiled in)
Titan · Projected Output · 200 VUs · 60s
══════════════════════════════════════════════
🚀 NexoLoad Titan — Test Summary
══════════════════════════════════════════════
URL lambda/prod/ping
Users 200
Engine asyncio + aiohttp
──────────────────────────────────────────────
Total Req 172,190
Successful 172,190 (100%)
TPS (engine) 2,824.0
TPS (success) 498.0 ← API GW cap
──────────────────────────────────────────────
Avg 61ms
P50 58ms
P95 83ms
P99 133ms
──────────────────────────────────────────────
Result: EXCELLENT
New Variant

Introducing Titan Raw

No dashboard. No colors. No spinner. Just the numbers — on one line. Built for pipelines that can't afford noise.

Titan Raw · --raw flag

Strip Everything. Keep the Truth.

Add --raw to any Titan command and the entire dashboard disappears. No progress bars. No live stats. No ANSI codes. When the test finishes, you get exactly two lines — everything a CI/CD gate needs to make a decision.

# Standard Titan run with --raw flag
$ ./nexoload-titan --url https://api.example.com --users 1000 --duration 60 --sla 400 --raw

... 60 seconds pass silently ...

NexoLoad Titan · 1000 VUs · 60s · asyncio
TPS: 5,504 | P95: 88ms | P99: 112ms | Errors: 0.00% | Req: 386,943 | SLA: PASS | Result: EXCELLENT
✓ CI/CD ready
✓ pipe-safe output
✓ zero ANSI codes
✓ grep-friendly
✓ non-TTY compatible
✓ same binary · just add --raw
Advanced Control

7 Flags That Go Further

Titan ships with all 18 Pro flags plus 7 advanced flags for edge cases Pro can't handle.

--target-rps N
Cap global TPS — throttle all VUs to stay at or below N requests per second. Useful for testing rate-limited APIs without overwhelming them.
--connections N
Cap max concurrent TCP connections. Useful when your server enforces connection limits independently from concurrent requests.
--warmup N
Discard all results during the first N seconds. Let the API warm up before the clock starts — keeps your P95 numbers honest.
--assert-status N
Treat any HTTP response other than status code N as a failure. Catch unexpected 200s on endpoints that should return 201 or 204.
--pacing-jitter N
Add 0 to N seconds of random delay per VU per request. Simulates realistic human traffic patterns instead of perfectly synchronized hammering.
--connect-timeout N
Set a separate TCP connection timeout independent from the total request timeout. Catch slow connection establishment without affecting read timeouts.
--no-verify-ssl
Disable SSL certificate verification. Test internal APIs behind self-signed certificates without any workarounds.
--raw
Titan Raw mode. No dashboard, no colors, no progress. Just two lines at the end with every key metric. Built for CI/CD pipelines and non-TTY environments.
Pick Your Edition

When Do You Need Titan?

💼 Pro

~200 VUs · OS thread engine

Perfect for everyday API validation, SLA monitoring, and CI/CD gates at normal concurrency. If your tests stay under 200 VUs, Pro is all you need.

🚀 Titan ← You Are Here

10,000+ VUs · async engine

The moment you need more than 200 concurrent users, Titan is the answer. Same binary size. Same flags. The thread engine simply can't go further — Titan can.

⚡ Helios

80,000+ VUs · multiprocessing (coming soon)

When Titan's single-core event loop hits its ceiling, Helios multiplies the engine across every CPU core. One worker process per core — targeting 80,000+ VUs.

First-Run Notice // When Titan Ships
⚠️
Titan macOS First-Run Notice

macOS Gatekeeper will block the unsigned Titan binary on first launch. Right-click the file and select Open, then confirm — or run the following once in Terminal:
xattr -rd com.apple.quarantine nexoload-titan-mac
One-time step. The binary runs normally on all subsequent launches.

⚠️
Titan Windows First-Run Notice

Windows SmartScreen will flag the unsigned Titan executable on first launch. Click More info, then Run anyway to proceed. One-time step caused by the binary not having a paid Microsoft code signature — not a security threat.

Titan Is Coming — Fall 2026

Titan is currently in pre-release. Contact us to schedule a live demo, discuss enterprise pricing, or get early access when it ships this Fall 2026.

Request Pre-Sales Demo
← Compare all NexoLoad editions