Tokalator

Count your tokens like beads on an abacus.

Real-time context budget monitoring for VS Code. Shows where your AI context budget is going, scores tab relevance, and syncs to the active Copilot model automatically — so you always know how much room you have left.

The Problem

AI coding assistants have finite context windows. When you have 30 tabs open, your assistant's attention is diluted across irrelevant files — and you can't see what's happening. Context rot sets in silently: the model's performance degrades as conversation turns accumulate stale context.

Tokalator makes this visible. Real BPE tokenizers (not guesses), per-file breakdowns, and automatic cleanup.

Features

1. Token Budget Dashboard

Sidebar panel showing your context usage at a glance:

• Budget level: LOW / MEDIUM / HIGH with token counts
• Next turn preview — cost estimate before you send
• Files ranked by relevance to your current task
• One-click tab cleanup
• Per-turn context growth chart
• Budget breakdown: files, system prompt, instructions, conversation, output reserve
• Session summary on activation (last model, turns, peak usage)
• High Contrast theme support

2. Model Auto-Sync

Tokalator detects the model active in the Copilot chat window and updates its context window, tokenizer, and rot threshold automatically. Switch models in Copilot and the budget numbers update on the next interaction.

3. Chat Commands (@tokalator)

11 commands covering the full context management workflow:

4. Tab Relevance Scoring

Each open tab is scored R ∈ [0, 1] based on:

Pinned and active files always score 1.0. Files below the threshold (default 0.3) are flagged as distractors.

5. Instruction File Scanner (/instructions)

Detects files injected into every prompt and shows their real token cost:

.github/copilot-instructions.md · CLAUDE.md · AGENTS.md · .cursorrules · *.instructions.md

6. MCP Server + CLI

Brings Tokalator into Claude Code and terminal workflows. See the MCP & CLI section.

Tokenizers

Supported Models (17 profiles)

Claude Opus 4.6 · Claude Opus 4.5 · Claude Sonnet 4.6 · Claude Sonnet 4.5 · Claude Sonnet 4 · Claude Haiku 4.5 · GPT-5.4 · GPT-5.4 Mini · GPT-5.2 Codex · GPT-5.1 Codex · GPT-4.1 · o3 · o4-mini · Gemini 3.1 Pro · Gemini 3 Pro · Gemini 3 Flash · Gemini 2.5 Pro

Usage

1. Sidebar — click the abacus icon in the Activity Bar
2. Chat — type @tokalator in Copilot Chat
3. Status bar — live token count in bottom-right (click to refresh)

Settings

Changelog

v3.1.3

• Fix: session logger receiving wrong object shape — all logged fields were undefined
• Fix: logOptimize called with wrong arguments (threshold passed as token count)
• Fix: missing await on model switch caused stale token budget in response
• Fix: dashboard listener memory leak — subscription now properly disposed
• Model list updated to current models (Claude 4.5, GPT-4.1 family, Gemini 2.5)
• Model auto-detection improved: vscode.lm.onDidChangeChatModels listener added
• Unknown model warning when switching to a model not in profiles
• Security: Next.js 16.2.1 (5 CVEs patched), hono, flatted

v3.1.2

• Auto-sync model from Copilot chat window
• Fix: stale token counts after rapid file switching
• Fix: duplicate tab entries in multi-root workspaces
• Fix: CLAUDE.md and AGENTS.md now counted in instruction scanner budget
• Fix: pin/unpin event propagation in dashboard

v3.1.1

• CLI terminal showcase, community session stats, MCP & CLI install section

v3.1.0

• MCP server for Claude Code (stdio transport)
• Standalone CLI: tokalator count, budget, preview, models
• 11 model profiles across Anthropic, OpenAI, and Google

Requirements

• VS Code 1.99+
• GitHub Copilot or similar AI extension (for chat features)

License

MIT