Cloud for agents
Production deploys for agent-built apps.
No Dockerfiles. No CI pipelines. No infrastructure config. Just code.
Claude Code
Cursor
Codex
Gemini CLI
Windsurf
VS Code
Kimi Code
Open Code
Antigravity
GooseHow you ship with INK
Pick the operating model. The deployment surface, workspace boundaries, and agent entry points stay in one place.
For developers
Push code. Or let your agent push it.
Either way, live in under a minute. No infra to touch.
Push to INK git or GitHub, or let your agent do it through the CLI, Skills, or MCP. INK takes the build from there: stack detected, service deployed, logs and metrics available, billed only while it runs.
Git push deploy
branch → URL
Auto stack detect
Next, Rails, FastAPI, Go +
Database templates
Postgres, Redis +
Persistent volumes
durable storage
Custom domains
auto-TLS
INK or GitHub repos
auto-redeploy
Per-minute billing
while running
CLI, Skills, MCP
Claude, Cursor, Codex
Your services
Deploy. Monitor. Ship.
Create services from GitHub repos or Docker images, then watch them come alive with real-time metrics and logs.
Create from source
GitHub repo or Docker image becomes a running service.
Inspect runtime state
The same surface exposes health, metrics, and logs.
Templates & Databases
One click. Or one prompt.
Deploy databases and full-stack templates from the canvas, or let your agent do it autonomously through the CLI or MCP.
Stateful services
Databases and persistent storage live beside app services.
Repeatable starts
Templates turn common stacks into a deployable starting point.
Observability
See everything. So can your agent.
Real-time metrics and structured logs, accessible from the dashboard or programmatically through the CLI, Skills, and MCP.
Human dashboard
Filter logs, inspect charts, and find runtime issues visually.
Agent access
Agents can read the same signals without leaving the coding flow.
Structured logs.
Zero noise.
Filter runtime and build logs by level, full-text search, or query syntax. Errors surface instantly, for you and your agent.
Filter by level
Errors, warnings, info and debug, one click
Full-text search
Query syntax like @level:error or plain text
Build & runtime
Both build output and runtime logs in one place
Real-time metrics.
Always watching.
CPU, memory, and network I/O, charted in real time with configurable ranges. Your agent reads the same data to self-diagnose and scale autonomously.
CPU, Memory & Network
Real-time resource charts with configurable ranges
Threshold lines
See limits at a glance on every chart
Agent-accessible
Agents read metrics through the CLI or MCP to self-diagnose
0.035vCPU
80MB
Pricing
Pay for what you use
No idle charges. You only pay per minute while your services are running. Your agent controls CPU and memory allocation and can adjust resources in real time.
$5 trial credit to start, no credit card required.
View full pricingMemory
$0.000161
per GB / min
CPU
$0.000393
per vCPU / min
Volumes
$0.0000035
per GB / min
Universal deployment
Deploy any stack
From static sites to full-stack apps, your agent deploys whatever it builds. No config files, no Docker setup, no CI pipelines, just code and go.
Docker, buildpacks, or auto-detected. Use one deploy surface across runtimes instead of choosing a different platform per stack.