Why I built Scopeful

I bought a month of Higgsfield last spring. The pricing page said Unlimited. It wasn’t. Priority queue is a separate credit pool that refills monthly and doesn’t roll over, buried two clicks into the fine print. By week three I was waiting 12 minutes per generation on a client deadline.

Nobody was going to fix this. Vendorswon’t, because confusion is their margin. Directories won’t, because they’re scraping logos and taking ad money. So I open each tool’s pricing page by hand, convert credits to dollars per output, and write the date on the row. Those numbers live here.

How every price gets verified

I open each tool’s pricing page by hand, compare totals against the database, and update anything that has shifted. A single pass checks plan prices, credit allotments, per-generation costs, access tiers, and resolution coverage.

Beyond pricing

Scopeful also maintains a growing library of agent skills, ready-to-use integrations for Claude Code, OpenCode, Windsurf, Cursor, Gemini, and Codex. Each skill is a self-updating module (SKILL.md) that gives your coding agent real-time knowledge of a specific AI tool's pricing, models, and API surface. Skills are free and public.

The MCP server exposes the full pricing engine over the Model Context Protocol. Any MCP-compatible agent can query costs, compare tools, and calculate workloads directly from your editor. Workflows and collections bundle skills and step-by-step tool chains for common creative pipelines (ComfyUI, multi-model video, image-to-video branching).

The free credits vault tracks over $700K in estimated credits across 50+ AI tool programs: free tiers, trials, startup grants, creator programs, and community challenges, with verified claim URLs and eligibility notes.