Now in Early Access

MekongMind
AI Model Router
as a Service

290 commands. Zero-cost inference. One API key.
Route developer missions to the optimal AI model automatically — Qwen, DeepSeek, Llama, or Claude — based on task complexity and cost.

Start Free See how it works
290
Commands
8
Pipeline stages
4+
AI models
$0
Inference cost

Built for developer teams
who move fast

MekongMind abstracts the complexity of multi-model AI orchestration into a single REST API.

Intelligent Model Routing

Automatically dispatches missions to the optimal model based on task type, complexity, and cost profile. Supports Qwen, DeepSeek, Llama, and Claude as fallback.

qwen · deepseek · llama · claude

MCU Credit System

Pay per mission execution, not per token. MCU (Mission Credit Unit) pricing gives predictable costs and aligns billing with business outcomes — not compute consumption.

mission-based billing

Multi-Tenant Isolation

Each workspace operates with its own API key and isolated execution context. No data leakage between tenants. Suitable for teams and enterprise deployments.

workspace-scoped keys

8-Stage Execution Pipeline

Every mission runs through a structured pipeline: classify, validate, plan, route, execute, review, format, deliver. Consistent output quality regardless of model used.

classify → route → execute → review

From API key to results
in under 60 seconds

MekongMind integrates with your existing workflow via REST. No SDK required.

01

Sign up & get API key

Create a workspace and receive your mk_ prefixed API key instantly.

02

Send mission via REST API

POST your goal as a plain-text instruction to the /raas/missions endpoint.

03

MekongMind routes to optimal model

The 8-stage pipeline classifies, plans, and dispatches to the best available model for your task.

04

Get results, credits deducted

Receive structured output. One MCU credit is deducted per successful mission execution.

One endpoint.
Any mission.

Simple REST interface

A single POST endpoint accepts any natural-language goal. No complex request schemas or SDK dependencies.

Bearer token authentication

Authenticate with your workspace API key. Keys are scoped per workspace and can be rotated at any time.

Structured JSON response

Responses include the result, model used, credits consumed, and pipeline trace for full observability.

bash
# Submit a mission to MekongMind
curl -X POST https://api.mekonmind.com/raas/missions \
  -H "Authorization: Bearer mk_your_key" \
  -H "Content-Type: application/json" \
  -d '{
    "goal": "Write a marketing pitch for
             Vietnamese coffee brand"
  }'

# Response
{
  "status": "completed",
  "result": "...",
  "model_used": "qwen-max",
  "credits_used": 1,
  "pipeline_ms": 1840
}

Transparent, mission-based pricing

Pay for what you execute. No per-token surprises. Unused credits roll over monthly.

Free
Starter
Try it free
$0/mo
50 MCU credits / month
50 MCU credits
REST API access
Community support
Get Started
Team
Growth
For growing teams
$149/mo
1,000 MCU credits / month
1,000 MCU credits
Up to 5 workspaces
Priority routing
Slack support
Get Started
Enterprise
Pro
For enterprises
$499/mo
5,000 MCU credits / month
5,000 MCU credits
Unlimited workspaces
SLA + dedicated support
Custom model integration
Get Started