whereismyai
Agent Smith
Agent Smith is an open source developer tool that routes tasks to smaller local models and monitors usage.
Agent Smith is an independent open source project designed to reduce token overspending during coding tasks. It consists of an adaptive router and a native macOS floating usage monitor called Codex Float. The adaptive router performs local model selection, user preferences, evidence-backed acceptance, and reproducible benchmarks using a local Codex model router. Codex Float runs as a native macOS floating window to track tasks, agents, usage, and reference costs. It writes a numeric SQLite ledger for deduplicating usage across restarts and log rotation. It targets developers seeking transparent model routing and precise usage tracking.
Agent Smith at a glance
Key features
- Local model selection
- Evidence-backed acceptance
- Reproducible benchmarks
- Native macOS floating window
- Numeric SQLite ledger
- Subagent dispatch and acceptance
Use it for
- Routing routine work to lighter models
- Monitoring token usage and reference costs
- Running offline benchmark demonstrations
- Inspecting child agents and routing evidence
Companions
Tools that articles about Agent Smith recommend using with it, each with its job.
- Context7 — Provides up-to-date documentation for AI agents and code editors.
- Ollama — Run local models at zero marginal cost, with code kept on your machine.
- OpenRouter — Access over 500 models and 80 providers through one unified API.
- Daytona — Runs sessions in isolated sandboxes synced to git branches.
- Oh My OpenCode Slim — Lightweight multi-agent orchestration and background tasks.
Alternatives to Agent Smith
- OpenCode — Works with 75+ model providers, local models, and existing subscriptions
- Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
- jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
- Open Agents Template — Spawn autonomous coding agents with full sandboxes and durable workflows.
- LangSmith — Step-by-step traces and evals that show what an agent actually did
- LangChain — Compose exactly the agent your use case needs from model, tools, prompt, and middleware.
- Agentfork — It replays steps from cache and runs sweeps to provide pass rates instead of anecdotes.
- Codex Infinity — Autonomous agents plan, test, and fix CI in isolated sandboxed environments.
- Weft — It enables safe parallel coding agent work without file locks.
- Agent Relay — It carries turns between different AI agents and stops for human decisions.
- AgentDock — Manages multiple coding agents from one screen-edge inbox without switching tabs.
- Command Code — Repairs open model tool calls and learns developer patterns automatically.
All Agent Smith alternatives