Flash was used to build Flash
Flash is an ultra-fast MoE built specifically for rapid task completion.
Up to 1200 tok/s.
Flash task preview
Local preview — nothing leaves your browser.Recorded task
Implement sixteen independent checkout modules against frozen contracts, then run one combined check.

One product. Every way you work.
Bring Flash to your harness, IDE, or API workflow and move from prompt to finished task faster.
Run Flash in Claude, Codex, Pi, Hermes, and compatible harnesses.
Use Flash inside the coding environment you already work in.
Call Flash through the OpenAI-compatible API for full control.

Your AI leads. Flash executes.
Keep planning, decisions, and final review with your preferred assistant. Delegate bounded tasks to Flash, running independent work in parallel when your harness supports it.
The executor and orchestrator setup guides configure your role and install this Flash skill; already-configured users can use it as a standalone skill from here. The coordinator can be Flash or another assistant.
View skill instructions

Bring Flash to your workflow.
Claude, Codex, Pi, Hermes, your IDE, or the API—Flash is ready wherever you start the task.
Choose how much work Flash can take on
Every plan includes Flash across harnesses, IDEs, and API workflows. Up to 1200 tok/s.
Premium
Complete focused tasks with Flash.
Features
Privacy-first processing
Prompts and responses are not used for training
Growth
Most popularComplete more work with Flash.
Features
5x usage
Privacy-first processing
Prompts and responses are not used for training
Scale
Highest capacityPut Flash behind demanding workflows.
Features
20x usage
Privacy-first processing
Prompts and responses are not used for training

Everything you need to know about Flash
Product, speed, compatibility, access, and API answers in one place.
Ready to work with Flash?
Join the waitlist to bring Flash into your workflow.

