Toki Labs

International traffic,
delivered compliant and on time

Toki Labs provides businesses with international IPLC private networking and one unified AI model API. Our first connection, Shanghai–Japan IPLC, is live with about 25 ms latency from mainland China to Japan and a free trial, and more connections are on the way.

Enterprise private networking, built to reach overseas

IPLCDedicated point-to-pointEnterprise grade

International private connection

Type
IPLC international private line
Coverage
Shanghai–Japan live, more planned
Billing
Dedicated bandwidth, six tiers from 10 to 100 Mbps
Terms
Monthly, quarterly, 6 months, 1, 2 or 3 years
Best for
Cross-border store operations, overseas service access, office interconnect
View connections

Low-latency direct path

Traffic takes a private path straight to the overseas node instead of detouring across the public internet, so latency stays low and steady.

Dedicated point-to-point

The link between both ends is yours alone, so peak-hour internet congestion doesn't slow you down.

Pick the bandwidth you need

Dedicated, unmetered bandwidth in six tiers from 10 to 100 Mbps. Move up a tier as your business grows.

24/7 support

Engineers monitor your connection around the clock and handle setup, migration and configuration directly.

Connections

Each connection is ordered and billed separately. Shanghai–Japan is live today, with a free trial.

Live

Shanghai–Japan IPLC

China Telecom Shanghai to Japan IPLC, about 25 ms from mainland China to Japan. Dedicated bandwidth from 10 to 100 Mbps.

Free trialOn-site hardware
See plans →
Planned

More international connections

We're planning new connections based on customer demand. Each one passes a compliance review before launch.

Request

Tell us where you need to connect

Share the destination and bandwidth you need, and we'll tell you as soon as it launches. Submit a request →

Free trial

Test with your real workload before you buy. Apply and we'll open a [trial period] Shanghai–Japan trial. Order only when you're happy with the speed and stability.

Request a free trial

On-site hardware

We supply a preconfigured device and install it at your office or data center. Plug it in, and every computer and device on your network uses the private connection with no per-device setup.

Ask about on-site hardware

Shanghai–Japan IPLC plans

Billed monthly · quarterly · 6 months · yearly · 2 years · 3 years

10Mbps
Dedicated · unmetered
¥[price] / month
Buy now
20Mbps
Dedicated · unmetered
¥[price] / month
Buy now
30Mbps
Dedicated · unmetered
¥[price] / month
Buy now
40Mbps
Dedicated · unmetered
¥[price] / month
Buy now
50Mbps
Dedicated · unmetered
¥[price] / month
Buy now
100Mbps
Dedicated · unmetered
¥[price] / month
Buy now
  • Every plan includes
  • Point-to-point IPLC link
  • Unmetered traffic
  • About 25 ms from mainland China to Japan
  • Free trial
  • Optional on-site hardware
IPLC

International private line, dedicated end to end

~25 ms

Latency, mainland China to Japan

7×24

Engineers on call

One API for a growing set of AI capabilities

One endpoint and one key, compatible with the OpenAI API format. Use it for coding and image generation today. When video generation and other capabilities arrive, they come through the same API and your code stays the same.

AI coding

Connect leading code models to IDE plugins, command-line tools and your own systems, so your team writes, reads and changes code faster.

  • Code generation and completion
  • Code review and refactoring
  • Debugging and writing tests
  • Works with common IDEs and CLI tools

Image generation

Generate images from text or a reference picture for product shots, marketing assets and concept art. Output carries AI-generated labels as regulations require.

  • Text to image
  • Edit and extend reference images
  • Product shots and marketing assets
  • Multiple sizes and aspect ratios

Video generation

Planned

Video and other new capabilities will join the same API. Once live, switch the model name to use them. Your integration and billing stay unchanged.

  • Same endpoint and key
  • One usage and billing view

OpenAI-compatible. Change one line to connect.

OpenAI-compatible

Keep your existing SDKs and tools. Just swap the base_url and key.

Keys per project

Give each project its own key and quota in the console.

Clear usage and billing

Coding is billed per token and images per picture. Every call shows up in your dashboard.

Content safety checks

Both prompts and outputs are screened. Requests that break the rules are blocked and logged.

python
from openai import OpenAI

client = OpenAI(
    base_url="https://api.tokilabs.xyz/v1",
    api_key="sk-toki-••••••••",
)

resp = client.chat.completions.create(
    model="[coding model]",
    messages=[{"role": "user",
               "content": "Write unit tests for this function"}],
)
# Use the same endpoint and key in your IDE plugin

We only offer models that can be lawfully provided in your region. See the console for the current list.

Compliance is our first product

International networking and generative AI are both regulated businesses. We run ours with clear service agreements, content safety controls and data protection, so your business can operate steadily for the long term.

Clear service agreements

Every customer signs a service agreement before activation that sets out permitted use, responsibilities and service levels.

Encryption and privacy

Links and account data are encrypted. We never sell or share customer data.

Lawful use only

Connections are for lawful business such as cross-border commerce and office connectivity. Use that breaches the agreement ends the service.

Logs kept as the law requires

We keep the logs that applicable law requires and disclose them only on lawful request.

Generated content review

Our AI service screens inputs and outputs and blocks unlawful or harmful content.

AI content labels

Generated images carry visible and embedded AI-generated labels so their origin is clear.

Contact

Start today on a connection that runs on time