Chinilla

What is Chinilla?

Chinilla is a system design simulator and architecture diagram tool for building and stress-testing service architectures visually. It helps you design systems using a component-based workflow, run deterministic simulations of traffic through your design, and identify where bottlenecks or failures occur.

Chinilla also includes an AI assistant that can read a simulation (and, separately, source code or specs) and explain what went wrong while walking you through fixes and design changes.

Key Features

• Visual system builder with drag-and-drop components: Add components and rewire flows to model how requests move through a system.
• Deterministic simulation with timeline inspection: Run traffic through your design and scrub the timeline to inspect a specific moment, including animated request flow and live statistics.
• 12 behavior modes for realistic failure and load patterns: Use modes such as queues, retry, and circuit breakers to represent common operational behaviors.
• Stress testing for spikes and outages: Apply traffic spikes and outages to see when queues fill, requests drop, and bottlenecks form.
• Universal modeling blocks for different domains: Use seven universal blocks intended to model systems across domains (examples given include backends, kitchens, hospitals, and factories), without requiring infrastructure jargon.
• Chinilla AI for critique and diagram generation: The AI reads your design/simulation to explain issues and propose fixes; it can also parse code or specs to generate an interactive diagram.
• Code/spec ingestion and editing: Paste code, drag and drop, or upload code files up to 100KB for diagram generation; support includes 15+ languages plus YAML, JSON, TOML, and XML configs.
• Exports for documentation and reuse: Export PNG, SVG, and animated GIF outputs; also export Python and Mermaid for code-level integration, plus a full project backup.

How to Use Chinilla

1. Start a new design using either a blank canvas or one of the provided templates.
2. Build the architecture visually by dragging components into place and rewiring the request flow.
3. Select behaviors and run a live simulation to observe how traffic moves through the system, then scrub the timeline and review live stats to find bottlenecks.
4. Use Chinilla AI to iterate: After running the simulation (or after providing code/specs), review the AI’s explanation of what went wrong and incorporate suggested changes.
5. Export outputs as needed (e.g., PNG/SVG/GIF for docs, or Python/Mermaid for further work).

Use Cases

• Practice system design interviews: Load an interview problem template, run the simulation, and check whether your design holds under the included metrics, behaviors, and costs.
• Debug performance bottlenecks before they happen: Simulate queue growth, request drops, and failure points by applying traffic spikes and outages.
• Turn existing code or configuration into an architecture diagram: Upload or paste a focused module or single file (up to 100KB) and use the generated diagram as a starting point to simulate and refine.
• Prepare documentation and presentations: Export static diagrams (PNG/SVG) or animated GIFs that show the system running, for READMEs and presentations.
• Create reusable interview-style templates and learn patterns: Use templates and step-by-step lessons to build systems repeatedly and apply AI critiques to improve future designs.

FAQ

• Is Chinilla available on desktop and mobile? The website states it is desktop only (for now).

• What file sizes and input types does Chinilla AI support? Chinilla AI supports uploading code files up to 100KB, and it can read 15+ languages plus configuration formats including YAML, JSON, TOML, and XML.

• What outputs can I export from Chinilla? You can export PNG, SVG, and animated GIF files for documentation, as well as Python and Mermaid for code-level integration, and you can keep a full project backup.

• Does the simulator help with failure scenarios? Yes. It includes stress testing with traffic spikes and outages, and it supports behavior modes such as queues, retry, and circuit breakers.

• How does Chinilla AI fit into the workflow? Chinilla AI can explain what went wrong in a design/simulation and walk you through fixes; it can also generate interactive diagrams by reading code or specs you provide.

Alternatives

• Architecture diagram tools (e.g., general diagramming software): Useful for drawing diagrams, but they typically don’t provide a built-in deterministic traffic simulation with timeline inspection and stress testing.
• System design interview prep platforms: Often focus on guided lessons and review, but may not include a visual simulator that runs traffic and shows where queues, drops, and bottlenecks appear.
• Code-based simulation/modeling tools: Can model complex systems and failures, but usually require more implementation effort and may not offer the same visual, component-based editing workflow.
• Whiteboarding tools with templates: Helpful for collaborating on diagrams, though they generally don’t parse code/specs to generate diagrams or run traffic simulations with live stats.