CrossGen

Based on the available evidence, CrossGen is better judged as an early open-source image workspace worth watching, not a broadly validated production tool yet. The main evidence is its GitHub repo description, which supports its positioning as local-first and covering generation, editing, organization, and reuse across a desktop app, JSON CLI, and MCP. That is enough to describe scope, but not enough to verify reliability, output quality, or real-world workflow performance.

In practice, it appears closer to a local image workspace that both humans and agents can call into, rather than a single-purpose text-to-image app. An easy misconception is to treat it as just another model or a cloud design SaaS. A more accurate analogy is a local-first image generation and asset workflow layer that combines desktop usage with programmatic access.

On barriers and cost, the evidence only supports that it is an open-source, local-first project. There is not enough support for claims about pricing, free usage, infrastructure requirements, or API costs.