Alpha · Orkait

Cloud runtime for untrusted code.

Execute code from users, agents, and LLMs in kernel-isolated sandboxes. 8 languages, 36ms latency, evidence-backed verdicts.

36ms

p50 latency

8langs

supported

3layers

kernel isolation

Languages supported

PythonJavaScriptTypeScriptCC++JavaGoRust

terminal

$ curl -X POST \
https://rustbox-api.orkait.com/api/submit \
-H "X-API-Key: rb_live_****" \
-d '{"language":"python","code":"print(42)"}'

ACAccepted

python · completedruntime / normal_exit

Output

stdout

stderr

empty

Timing

CPU time9 ms

Wall time10 ms

CPU/wall0.900

Patterncpu bound

Process

Exit code0

Signalnull

Procs0 / 10

Memory

Peak3.68 MB

Limit256.00 MB

OOM events0

OOM kills0

cgroup CPU9.1 ms

Security

isolationstrict

pid namespacemount namespacenetwork namespacememory limitprocess limitno new privileges

Missingnone

Containmentok

Reapclean

Zombies0

Metadata

Schema1.0

Integritycomplete

Created2026-04-03 08:05:23Z

Started2026-04-03 08:05:23Z

Completed2026-04-03 08:05:23Z

id: f2c6c5f7-1203-46e5-8a4f-a619c12bfeb0

Have a key? Open playground

Product

Two profiles. Judge mode live now.

One API, two isolation profiles. Your API key maps to a profile - same endpoint, same contract, different resource limits.

Assessment Platform

[2026]Compiling solution...

#include <iostream>
using namespace std;

int main() {
  int n; cin >> n;
  // O(N) solution
  long long sum = 0;
  for(int i=1; i<=n; i++)
    sum += i;
  cout << sum;
  return 0;
}

Rustbox Kernel

ISO-LEVEL: 4

Running tests...

LIMIT

4s CPU

RAM

256 MB

NET

Off

Execution Verdict

Test Suite Execution

Test #112ms

Test #218ms

Test #3Wrong Answer

Judge

Competitive programming, coding assessments, automated grading.

Open Beta

Memory

256 MB

Wall

CPU

Procs

No network, no loopback, no DNS

Executor

LLM tool execution, interactive environments, long-running tasks.

On Request

Memory

2 GB

Wall

60s

CPU

30s

Procs

Outbound-only via nftables

8 languages supported

Python
C
C++
Java
JavaScript
TypeScript
Go
Rust

Playground

Run code before wiring the API.

Try Judge mode from the landing page: switch languages, edit code, resize panels, provide stdin, and inspect the same structured verdict shape your API integration receives.

Loading playground...

Security

Why trust Rustbox with untrusted code.

Every execution passes through 8 kernel isolation layers. Order enforced at compile time by Rust's type system.

8 kernel isolation layers

PID namespace, mount namespace, network namespace, cgroup v2, seccomp-bpf (52 rules), no_new_privileges, read-only rootfs.

Compile-time enforced

Rust typestate chain. Skip a step and it does not compile. Order is a language guarantee, not a runtime check.

Tested against real attacks

22 adversarial vectors across all 8 languages. Fork bombs, symlink escapes, /proc abuse, ptrace - all blocked.

escapes

across 22 adversarial attack vectors

440+

tests

vectors

recovery

Every API call gets the same isolation.
No tiers. No shortcuts.

Performance

Real numbers. No asterisks.

Sieve of Eratosthenes (n=500,000), full kernel isolation, single-threaded.

Sequential latencyp50 ms

Python

36 ms

JavaScript

36 ms

TypeScript

36 ms

60 ms

C++

72 ms

Java

445 ms

Throughput4 workers, req/s

Python

148/s

JavaScript

104/s

TypeScript

83/s

77/s

C++

70/s

Java

5/s

AMD Ryzen 7 9700X · 4 vCPUs · 4 GB RAM · Linux 6.12

API

First result in 60 seconds.

One POST, one structured response. No SDK, no WebSocket, no polling for simple submissions.

Open Playground API Reference

Get your API key

X-API-Key: rb_live_················

Make your first call

curl -X POST https://rustbox-api.orkait.com/api/submit \
  -H "Content-Type: application/json" \
  -H "X-API-Key: rb_live_your_key_here" \
  -d '{"language":"python","code":"print(42)","stdin":""}'

Read the structured verdict

{
  "completed_at": "2026-04-03T08:05:23.570195915+00:00",
  "cpu_time_secs": 0.009142,
  "created_at": "2026-04-03T08:05:23.540110054+00:00",
  "error_message": null,
  "evidence": {
    "cgroup": {
      "cpu_usage_usec": 9142,
      "memory_limit_bytes": 268435456,
      "memory_peak_bytes": 3862528,
      "oom_events": 0,
      "oom_kill_events": 0,
      "process_count": 0,
      "process_limit": 10
    },
    "collection_errors": [],
    "controls_applied": [
      "pid_namespace",
      "mount_namespace",
      "network_namespace",
      "memory_limit",
      "process_limit",
      "no_new_privileges"
    ],
    "controls_missing": [],
    "isolation_mode": "strict",
    "judge_actions": [],
    "process_lifecycle": {
      "descendant_containment": "ok",
      "reap_status": "clean",
      "zombie_count": 0
    },
    "timing": {
      "cpu_ms": 9,
      "cpu_wall_ratio": 0.9,
      "divergence": "cpu_bound",
      "wall_ms": 10
    },
    "verdict_actor": "runtime",
    "verdict_cause": "normal_exit"
  },
  "exit_code": 0,
  "id": "f2c6c5f7-1203-46e5-8a4f-a619c12bfeb0",
  "job_status": "completed",
  "language": "python",
  "memory_peak_bytes": 3862528,
  "output_integrity": "complete",
  "schema_version": "1.0",
  "signal": null,
  "started_at": "2026-04-03T08:05:23.540218508+00:00",
  "stderr": "",
  "stdout": "42\n",
  "verdict": "AC",
  "wall_time_secs": 0.01
}

Verdict reference

ACAccepted

RERuntime Error

TLETime Limit

MLEMemory Limit

SIGSignaled

IEInternal Error

Get your API key

1,000 submissions · 30 days · no credit card

1,000

submissions

148

req/s peak

languages

30 day

validity

Pricing

Agent profile

Higher-room sandboxing for tool calls, interpreters, and short-lived REPLs.

waitlist

filtered network1 GB memory30s wall

Networkfiltered

Queuetool-call ready

Runtimeextended wall

Free

Evaluate

20/day

Safe public cap

Included volume

20/day

Try the Agent profile at a safe public cap.

2 requests / minute
20 requests / day
1 GB memory - 30s wall
Filtered network
All 8 languages
Community support

Start free

Pro

Scale

Unlimited

Priority agent route

$49/month

Included volume

Unlimited

AI agent tool-calls without the stingy cap.

20 requests / minute
Unlimited requests / day
1 GB memory - 30s wall
Filtered network
Priority queue
Email support (48h)

Join waitlist

Business

Enterprise

Unlimited

Custom policy lane

$299/month

Included volume

Unlimited

Production agent workloads with SLA.

100 requests / minute
Unlimited requests / day
Custom network policy
Dedicated queue
99.9% SLA
Slack support (8h)

Contact sales

FAQ

Questions before you wire it in.

Short answers for the constraints that usually matter before teams add sandbox execution to a product.

Need a specific answer?

Send the workload shape, expected volume, language mix, and security requirements through the contact page.

Cloud runtime for untrusted code.

Two profiles. Judge mode live now.

Run code before wiring the API.

Why trust Rustbox with untrusted code.

8 kernel isolation layers

Compile-time enforced

Tested against real attacks

Real numbers. No asterisks.

First result in 60 seconds.

Get your API key

Make your first call

Read the structured verdict

Verdict reference

Get your API key

Agent profile

Free

Pro

Business

Questions before you wire it in.

What is Rustbox?

Which languages are supported?

Is network access allowed?

How does billing work?

Can I use Rustbox for AI agents?

How do I get support or request access?