Dime
The token-frugal AI coding agent for VS Code. Bring your own OpenAI, Claude, or Gemini API keys — Dime makes every token count and shows you the receipt.
Why pay $100–$200/month for AI coding subscriptions? Dime delivers agent-grade coding help on your own pay-as-you-go keys, engineered so a typical heavy user spends $10–$25/month — and it proves it with a cost receipt after every task.
✔ Task receipt
models Claude Haiku 4.5 (search) · Claude Sonnet 5 (edits)
tokens 18.2K in (14.1K cached) · 2.4K out
cost $0.026 saved vs. unoptimized: ~$0.31 (92%)
today $0.41 of $1.00 daily budget
Why Dime costs less
- Surgical context — long files are sent as structure outlines (symbols + line ranges, ~hundreds of tokens) instead of full contents (often tens of thousands). The agent reads only the exact line ranges it needs.
- Smart model routing — searches, tests, and simple jobs go to economy models (fractions of a cent); flagship models are reserved for architecture and hard debugging. Pin any model manually whenever you want.
- Prompt caching — conversation history is cache-structured so providers re-read it at up to ~90% discount instead of full price.
- Diff-only edits — the agent returns targeted search/replace changes, never wasteful full-file rewrites.
- Hard budgets — a pre-set daily spending ceiling (default $1.00) that actually pauses the agent, plus a per-task step cap that stops runaway loops.
Features
- Three providers, your keys — OpenAI, Anthropic (Claude), and Google (Gemini). Keys are stored in your operating system's secure keychain, never in files or logs.
- Automatic provider failover — if one provider runs out of credits or goes down mid-task, Dime switches to another provider you have a key for and carries the whole conversation over. No context lost, no re-explaining.
- Two editing modes — Approval (default): every file change is shown as a diff you accept or reject. Auto: edits apply immediately, with a checkpoint saved before every task and one-click rollback.
- Skills — Coding, QA/Tests, Security review, and Docs built in. Create your own from the chat dropdown (
＋ New skill…), or drop folder-based skills in ~/.claude/skills/, ~/.dime/skills/, or <project>/.dime/skills/ — any SKILL.md with YAML frontmatter is picked up automatically, including skills you already wrote for Claude Code.
- Clarifying questions — a vague request gets one short question with clickable options before any tokens are burned on the wrong interpretation.
- Attachments — attach text files (specs, requirements, logs) or images, or just paste a screenshot with Ctrl+V — with a token-cost preview before you send.
- Persistent chat history per project, saved as a plain JSON file you own at
~/.dime/history/ — survives reinstalls and upgrades, and is trivial to back up. Plus markdown-rendered replies and a live spend counter in the status bar.
- Terminal commands with consent — every command is shown and approved before it runs; destructive commands are blocked outright.
Getting started
- Install Dime and click its icon in the activity bar.
- Click Set API Key… and paste a key for at least one provider (all three work together, but one is enough):
- Open a project folder and ask for something: "explain what this project does", "add input validation to the upload handler", "write tests for utils.py".
Tip: also set a monthly spend cap in each provider's own billing console. Dime's daily budget is the first line of defense; the provider-side cap is the absolute stop.
Settings
| Setting |
Default |
What it does |
dime.mode |
approval |
approval reviews every change as a diff; auto applies immediately (checkpointed) |
dime.dailyBudgetUsd |
1.0 |
Hard daily spending ceiling — the agent pauses at this amount |
dime.maxStepsPerTask |
12 |
Maximum model calls per task |
dime.protectedGlobs |
.env, keys… |
Files the agent may never read, send, or edit |
dime.customSkills |
[] |
Your own skills (easier: use ＋ New skill… in the dropdown) |
Privacy & security
- 100% client-side: no Dime server, no account, no telemetry of any kind.
- Code goes only to the AI provider you chose, using your own key, over their official SDKs.
- API keys live in the OS keychain (VS Code SecretStorage).
- Protected files (
.env, private keys, certificates) are never read, sent, or edited; text matching common secret patterns is redacted from everything sent to a model.
- Project rules: add an
AGENTS.md file to your repo and the agent follows it.
Roadmap
- PDF attachments (images already supported)
- MCP (Model Context Protocol) tool support
- Remotely updated model registry — new models and prices without an extension update
- Command allowlist for Auto mode
- Local models (Ollama-compatible endpoints)
License
MIT — see LICENSE.
| |
