Intelligence Suite
EN

Models and Token Marketplace

One catalogueevery model.

One API across 100+ models from multiple providers, plus the keys, dashboards and controls to run AI consumption with confidence.

Open the AI Token MarketplaceSee pricing

How models are deployed

Three tiers. One perimeter.

All products

In-country Perimeter · Indonesia

Tier 01

Open source, locally hosted

Open-weight models on Indosat GPUs.

  • Sahabat-AI
  • GLM
  • Qwen
  • DeepSeek

Data stays in the nominated jurisdiction

Tier 02

Closed source, locally hosted

Licensed commercial models, hosted in-market.

  • Licensed
  • In-market

Residency set by policy per workload

Tier 03

Connected global models

Frontier models reached via the provider.

  • Anthropic
  • Google
  • xAI

The Catalogue

Every model, one connection.

  • Sahabat AIIn-country
  • Claude Haiku 4.5Anthropic
  • Claude Opus 4.5Anthropic
  • Claude Opus 4.6Anthropic
  • Claude Opus 4.7Anthropic
  • Claude Opus 4.8Anthropic
  • Claude Opus 5Anthropic
  • Claude Sonnet 4.5Anthropic
  • Claude Sonnet 4.6Anthropic
  • Claude Sonnet 5Anthropic
  • Deepseek v4 FlashDeepSeek
  • DeepSeek-R1 (0528)DeepSeek
  • DeepSeek-V3.2DeepSeek
  • Gemini 2.5 Flash (audio)Google
  • Gemini 2.5 Flash (text)Google
  • Gemini 2.5 Flash ImageGoogle
  • Gemini 2.5 Flash-Lite (audio)Google
  • Gemini 2.5 Flash-Lite (text)Google
  • Gemini 2.5 ProGoogle
  • Gemini 3 Flash Preview (audio)Google
  • Gemini 3 Flash Preview (text)Google
  • Gemini 3 Pro Image (preview)Google
  • Gemini 3.1 Flash ImageGoogle
  • Gemini 3.1 Flash-Lite (audio)Google
  • Gemini 3.1 Flash-Lite (text)Google
  • Gemini 3.1 Flash-Lite PreviewGoogle
  • Gemini 3.1 Pro PreviewGoogle
  • Gemini 3.5 FlashGoogle
  • Gemini 3.5 Flash-LiteGoogle
  • Gemini 3.6 FlashGoogle
  • Gemini 3.7 FlashGoogle
  • Gemini 3.8 FlashGoogle
  • Llama 4 MaverickMeta
  • Llama 4 ScoutMeta
  • MiniMax-M2MiniMax
  • Codestral 2Mistral AI
  • Mistral Medium 3Mistral AI
  • Mistral Small 3.1 (24B)Mistral AI
  • Kimi-K2-ThinkingMoonshot
  • gpt-oss-120bOpenAI
  • gpt-oss-20bOpenAI
  • Qwen3-235B-A22B-InstructQwen
  • Qwen3-Coder-480B-A35BQwen
  • Qwen3-Next-80B-InstructQwen
  • Qwen3-Next-80B-ThinkingQwen
  • Grok 4.1 Fast Non-ReasoningxAI
  • Grok 4.1 Fast ReasoningxAI
  • Grok 4.20 Non-ReasoningxAI
  • Grok 4.20 ReasoningxAI
  • Grok 4.3xAI
  • Grok 4.6xAI
  • GLM 5.3Z.ai
  • GLM-4.7Z.ai
  • GLM-5Z.ai
  • GLM-5.2Z.ai

Inside the marketplace

Build, monitor and govern in one place.

Illustrative screens with sample values

Model Catalogue

Switch models in one line

One OpenAI-compatible endpoint. Change the model, keep the code.

Request body diff: switching the model
curl $BASE_URL/v1/chat/completions -d '{
    "model": "sahabat-ai",    "messages": [...]
  }'
Switch models in one line
ModelInputOutputCache read
Sahabat AI0.100.37-
Claude Sonnet 52.0010.000.20
DeepSeek V4 Flash0.561.680.056

API Keys

Self-service, scoped keys

Create a key, limit its models, set an expiry, revoke instantly.

customer-care-botActive

sk-...kL3h

Models in scope

Expires
30 Sep 2026
Spend
$18.42
Last used
2 min ago
first call in minutes
first call in minutes

Dashboard

See where every token goes

Input, output and cached tokens over time, by model and period.

InputOutputCachedLast 4 weeks
Requests OK
128,406
Failed
317
Top model
Sahabat AI

Spend Controls

Hard ceilings on spend

Account-wide and per-key budgets, tokens and requests per minute.

Account budget$350 / $500

Budget resets

Budget resets on the 1st

Per-Key LimitsWithin the account limits

  • customer-care-bot
  • fraud-analytics
  • agent-sandbox

One runaway app hits its own ceiling, not the whole account.

Access every model through one connection.

Book a demoOpen the Token Marketplace