tokenwise
Measurement-driven model router for Claude Code. Routes Haiku/Sonnet/Opus per task class, logs every routed task with real $ numbers, and A/B tests cheaper tiers before you trust the savings.
IT & Software Developmentab-testinganthropicclaude-codecost-reductionhaikumeasurementmodel-routingopus
TokenWise — Measurement-Driven Model Router
Overview
A Claude Code skill that auto-routes subtasks to the cheapest model that can handle them (Haiku for grunt work, Sonnet for scoped reasoning, Opus only for synthesis), then logs every routed task to a local NDJSON with real token + cost numbers. Includes an A/B test subcommand that runs the same task across multiple tiers and scores quality, so the routing decisions are verified against the user's real workload — not estimated.
Anthropic's own bug tracker (Issue #27665) reports 93.8% of Max-subscriber Claude Code tokens flow to Opus. Existing routers (claude-router, wshobson, VoltAgent) either pin models statically or route by vibes-based heuristics with no measurement. TokenWise fills the measurement gap.
When to use
- Cutting Claude Code token spend without sacrificing output quality
- Validating whether Haiku/Sonnet is "good enough" for a specific task class before trusting auto-routing
- Auditing where Opus tokens are actually being burned
- Logging per-session cost data for finance or chargeback
Subcommands
/tokenwise:install— guided installer with diff preview, automatic backups, and--dry-runmode/tokenwise:report— per-session token + cost summary vs all-Opus baseline/tokenwise:summary [--week|--month|--all]— historical aggregate with trend/tokenwise:ab "<task>"— A/B test the
Subscribers only
The full skill, its 1 bundled files and every download is included with every paid Complete AI plan.
Details
| Source | CodeShuX/tokenwise |
|---|---|
| License | MIT |
| Risk label | critical ("critical" means the skill may run commands or touch files — read before use) |
| Files | SKILL.md |
| Added | 2026-05-12 |
Related skills
agents-generator
Generate project-specific AGENTS.md and companion rules by analyzing a codebase. Supports full, minimal, update, and dry-run modes with package-manager detection, monorepos, backups, managed blocks, confidence scoring, and command validation.
gh-attach
Upload and download GitHub user-attachments (screenshots, PDFs, zips, videos) from the terminal; use when asked to attach or embed a file in a PR, issue, or comment, or download an attachment URL.
gh-image
Upload local images to GitHub and get canonical user-attachments embed URLs; use when asked to attach a screenshot to a PR, issue, or comment, or to embed before/after images in a README.
mcp-tool-developer
Build Model Context Protocol (MCP) servers and tools from scratch. Full-stack MCP development with TypeScript/Python, testing, deployment, and registry publishing.