Beta Pricing — Cost Recovery Only

Pay for compute.Nothing else.

During beta, you pay exactly what the server costs us — no markup, no seat fees, no AI charges. Organizations and team members are always free.

Beta

$13.99/mo per agent

Raw compute cost — $0 Jarble margin

3 shared vCPU

4 GB

80 GB

Isolated pod with dedicated resources
Web chat interface with full canvas
30+ built-in components (charts, tables, code, 3D)
Connect WhatsApp, Discord, Slack, Telegram
MCP tool connections
Skills & component marketplace
Agent orchestration (flow engine)
Subagent delegation
Persistent storage across restarts
Real-time SSE streaming
Two-way config sync

Always free

Create unlimited organizations

Invite unlimited team members

Browse the marketplace

Access any LLM model via OpenRouter

AI Model Access

Compute and AI are billed separately — you control both

Bring Your Own Key

Recommended

Use your own OpenRouter, OpenAI, Anthropic, or Google key

Free— you pay your provider directly

Access to all models (GPT-4, Claude, Llama, Gemini)
Full control over model selection and costs
No markup on API costs — ever

Managed Credits

Optional

We handle the LLM key — just deploy and go

$5 – $100/mo credit cap

No API key needed — we provision one for you
Set a monthly spending cap ($5 – $100)
Same models, managed by Jarble via OpenRouter

Need dedicated infrastructure?

Custom compute, dedicated CPU, SSO, audit logs, or SLAs — let's talk.

FAQ

Common Questions

Ready to deploy your first agent?

Get started in minutes. No credit card required to explore.