Universal component system
Build diagrams from 7 universal component types: Person, Step, Storage, Decision, Trigger, Tool, and Channel. The small set is intended to stay composable while keeping the canvas readable.
Chinilla
Chinilla is a browser-based system design simulator for building and testing architectures on a canvas. Spot bottlenecks, queueing, and failure modes before code.
Overview
Chinilla is a browser-based system design simulator for mapping workflows, architectures, and other systems with moving parts. Users place components on a canvas, wire them together, and run a deterministic simulation to see how packets flow, where queues build, and where the design breaks under load.
The product is aimed at system design interview prep, architecture brainstorming, teaching, and early-stage validation before code is written. It also supports AI-assisted refinement, collaborative editing, and repository-to-diagram workflows for bringing existing projects into the canvas.
Features
Runtime behaviors and queueing
Connect nodes with behaviors such as passthrough, filter, split, delay, retry, circuit breaker, batch, and replicate. Queueing is also universal when a component has capacity set.
Deterministic simulation engine
Run a deterministic discrete-event simulation that pushes packets through the graph step by step. Same inputs and same seed produce the same result, which makes comparisons repeatable.
Load and failure analysis
Inspect where designs stall under load, including queue growth, overflow, drop rates, retry counts, rate-limit drops, and circuit-breaker trips. The runtime is meant to surface bottlenecks before implementation.
Collaborative interview workflow
Work on a shared canvas with live cursors, use interview mode for a mock round, and follow built-in rubrics. The site also describes one-on-one sessions with an owner and one guest as a Pro feature.
Import, publish, and export
Import from a GitHub repo or URL, then export results as PNG, SVG, Mermaid, animated GIF on Pro, or a one-click markdown System Spec. You can also publish a live link and embed the diagram in docs or a README.
Use Cases
Interview preparation
Load a system design prompt, build the architecture on the canvas, and run the simulator to see whether throughput, queueing, or failure behavior holds up under pressure.
Architecture brainstorming
Sketch microservices, workflows, or service dependencies before implementation to reason about capacity, bottlenecks, retries, and failure modes.
Teaching and learning
Use the canvas in lessons, docs, or guided demos to show how queues, delays, circuit breakers, and branching affect a system over time.
Repository-to-diagram workflows
Import an existing repository into Chinilla, turn it into a runnable diagram, and publish or embed the result for others to review.
Mock interviews and collaboration
Open interview mode with a built-in rubric for a mock round, then use the shared canvas and live cursors for collaborative sessions.
Pros and Cons
Pros
- Deterministic runs make results reproducible for debugging and before/after comparisons.
- The canvas uses a small set of universal primitives, which keeps the model compact while still allowing composition.
- It exposes runtime behaviors such as retries, circuit breakers, batching, and queueing, so the diagram reflects execution behavior, not just layout.
- It supports practical outputs and sharing, including live links, embeds, and common export formats.
- The pricing page shows a free option and a Pro tier with added AI design and GitHub repo support.
Cons
- It is not intended for production load testing or observability workflows.
- The site says it does not model wire-level network physics, jitter distributions, GC pauses, or kernel scheduling.
- The tool is scoped to design and learning rather than large industrial agent-based modeling or cloud deployment views.
FAQ
What is Chinilla used for?
Chinilla is a browser-based system design simulator. It is meant for architecture brainstorming, interview prep, teaching, and stress-testing designs before writing code.
Does Chinilla run in the browser?
The source describes Chinilla as a browser-based product, so the workflow runs in the browser rather than as a desktop install.
Is there a free version?
Chinilla’s pricing page shows a Free plan and a Pro plan at $20/month. The site also says the free version is available without a credit card.
What does the Pro plan add?
The site says Pro adds AI design and GitHub repo-to-diagram support. It also offers publishing, embedding, and export options such as PNG, SVG, Mermaid, and a markdown system spec.
What is Chinilla not meant for?
Chinilla is not positioned as a production load tester, observability tool, Kubernetes view, or wire-level network simulator. The site also says it does not model network jitter distributions, GC pauses, or kernel scheduling.
Quick Facts
- Category
- System design simulator
- Platform
- Browser-based
- Primary use cases
- Interview prep, architecture brainstorming, teaching, early-stage design validation
- Core model
- Deterministic discrete-event simulation
- Notable outputs
- PNG, SVG, Mermaid, markdown system spec, live links, embeds
- Pricing
- Free plan available; Pro is listed at $20/month
Chinilla Alternatives
HTTPie AI
HTTPie AI is an upcoming AI-powered API interaction tool from HTTPie, with Chat coming soon for a new way to work with APIs.
ClawTick
ClawTick is an AI agent automation platform for scheduling jobs from the CLI, dashboard, or REST API. Built for developers and teams using LangChain, CrewAI, webhooks, or custom workflows with monitoring, alerts, and logs.
Falconer
Falconer is an AI-powered company brain that keeps internal docs accurate, searchable, and up to date by syncing with GitHub, Slack, and Linear.
OpenFlags
OpenFlags is an open-source, self-hosted feature flag platform for JavaScript teams, with local evaluation, targeted rollouts, and controlled launches.
Speculos
Speculos deploys Claude Code apps to shareable URLs, with private, org-wide or public visibility and team controls for domains, data and cloud deployments.
Skills Janitor
Skills Janitor is a GitHub-hosted set of slash commands for auditing, tracking, and cleaning up Claude Code and OpenAI Codex skills.