← Back to home

AIDIA Token Hub

PLATFORM

TokenHub — one key to every model

One contract, one key, one invoice instead of a separate subscription with every model provider. The platform is already running in production.

Platform AIDIA TokenHub

TokenHub gives you access to language and multimodal models through a single OpenAI-compatible endpoint. Requests run through a node inside Kazakhstan, usage is metered by tokens, and every call is written to a log. Pay up front or under a corporate contract with an SLA.

Open TokenHub

Sign up, get a key, and send your first request without talking to a sales rep.

Migrating existing code — just change the base URL
# One key to every model: only the base URL changes
curl https://tokenhub.aidia.kz/v1/chat/completions \
  -H "Authorization: Bearer $AIDIA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "model-id",
    "messages": [{"role": "user", "content": "Hello!"}],
    "stream": true
  }'

OpenAI API compatibility

A REST API that follows the Chat Completions spec, SSE streaming, and support for the common SDKs. An existing integration moves over by changing the base URL.

Routing and failover

Load balancing across several channels, automatic failover to a backup channel when an upstream goes down, and automated health checks with recovery.

Billing and documents

Precise token-level metering, prepayment and pricing packages, promo codes, and low-balance alerts. Billing in tenge and US dollars with local invoices.

Keys, quotas, and access

Multiple users and keys with separated permissions, quotas and limits, QPS throttling, and IP allowlists — access control stays with your team.

Call log and audit trail

A complete request log and audit trail for compliance scenarios under the Law on Artificial Intelligence. Prompts and responses are not used for model training by default.

Usage monitoring

A spending dashboard broken down by key, model, and period, cost analysis, alerts on abnormal usage, and report and billing exports for reconciliation.

Protocol
OpenAI Chat Completions, SSE
Channel health checks
automatic, ~30 seconds
Processing node
inside Kazakhstan
Support
first line RU/KK, second line 24/7
Platform documents: Public Offer Privacy Policy
Discuss your task TokenHub