Token Cost Tracker for Copilot & Ollama Cloud

Fork Notice: This extension is a fork from https://github.com/assaelaz/copilot-cost-token-tracker, extended with Ollama Cloud model and pricing support, live pricing scraping from both GitHub Copilot and Ollama Cloud, configurable monthly AI Credit plans, and workspace-level filtering.

Visualise GitHub Copilot and Ollama Cloud agent session token costs directly inside VS Code. The extension runs locally, reads agent debug logs and chat sessions from your workspace storage and ~/.copilot/session-state, and shows the latest session, a full session table with workspace filters, an aggregate summary, and a dedicated Models pricing catalog.

Features

Dashboard views

• Latest Session — live per-model and per-provider breakdown of your most recent Copilot / Ollama Cloud agent session; auto-refreshes while the session is active
• All Sessions — sortable table of every session in the selected window with interactive Workspace toggle filters (All / None / individual workspaces), Cost, Credits, and Cache Hit % columns; click any row to expand per-model detail inline
• Aggregate — rolled-up totals across all sessions in the selected time window, broken down by model and provider (GitHub Copilot vs. Ollama Cloud)
• Models — unified catalog of known models across GitHub Copilot and Ollama Cloud, live remote pricing scraping (Update now), configurable AI Credit plan allowances (e.g., Ollama Cloud Pro $60 monthly credits for the $20/mo plan), and observed unmatched model tracking

Cost & token details

• USD cost and AI Credits (1 credit = $0.01) totals for every session and aggregate view
• Automatic provider disambiguation (GitHub Copilot vs. Ollama Cloud) even for overlapping model families (e.g., kimi-k3 / kimi-k2.7-code)
• Per-model token breakdown: fresh input, cached input, cache writes, and output tokens
• Cache hit rate per model with colour-coded health indicators
• Sub-agent call detection — badges distinguish main-agent vs. sub-agent LLM calls within a session

Time filtering

• Preset time windows: 1 h, 6 h, 12 h, 24 h, 48 h, 72 h, 7 days, 30 days, 90 days, and 365 days
• Custom date-range picker to query any arbitrary period

Pricing & model coverage

• Bundled media/pricing.json pricing table covering GitHub Copilot and Ollama Cloud model families
• Live pricing refresh from docs.github.com/en/copilot/reference/copilot-billing/models-and-pricing and ollama.com/pricing
• Intelligent model-key matching (provider prefix, :cloud tag, and substring) so minor version variants resolve to the exact provider and pricing rate
• Unknown models flagged with a warning and excluded from totals rather than silently misreported

Privacy & portability

• Reads only local Copilot / Ollama Cloud agent logs from VS Code / VS Code Insiders workspace storage and ~/.copilot/session-state
• Cross-platform support: macOS, Windows, and Linux storage paths resolved automatically
• No user session data leaves your machine

Requirements

• VS Code 1.85 or later, or VS Code Insiders
• Existing Copilot / Ollama Cloud agent session logs in your local workspace storage or ~/.copilot/session-state

Installation

From a VSIX

1. Download or build the .vsix file (npm run package)
2. Open the Extensions view in VS Code
3. Select ... and choose Install from VSIX...
4. Pick the .vsix file
5. Reload when prompted

Usage

Open the dashboard with any of these entry points:

• Status bar: Token Cost
• Activity bar: Token Cost Tracker (Ollama Cloud)
• Command Palette: Token Cost Tracker for Copilot & Ollama Cloud: Open Dashboard

Commands

Privacy

• Reads only local Copilot and Ollama Cloud agent session logs.
• Does not send session data or prompts to any external service (network requests are only made to GitHub Docs and Ollama pricing pages when refreshing model pricing).
• Unknown models are excluded from totals and flagged in the dashboard.

How It Works

The extension scans local VS Code workspace storage directories (debug-logs and chatSessions) as well as ~/.copilot/session-state, parses the JSONL session data, resolves each model's provider (GitHub Copilot or Ollama Cloud), and computes token usage and cost totals locally.

Model Coverage

The bundled pricing table (media/pricing.json) covers both GitHub Copilot and Ollama Cloud model families and can be updated on demand from the Models tab. If a model is not present in the pricing catalog, it is shown with a warning and excluded from totals.

Troubleshooting

• If the dashboard is empty, make sure you have an existing Copilot / Ollama Cloud agent session within the selected time window.
• If a model is missing pricing data, click Update now in the Models tab or check the Observed models without an exact price table.

Development

Prerequisites

• Node.js (v20+ recommended) and npm
• VS Code (^1.85.0) or VS Code Insiders

1. Install Dependencies & Build

Clone the repository, install dependencies, and compile the TypeScript sources (src/*.ts -> out/*.js):

npm install
npm run compile

To continuously recompile TypeScript files on save during development:

npm run watch

2. Verify Core Runtime Logic (CLI Test)

After compiling, you can verify pricing resolution, session discovery, and token/cost parsing directly from the terminal without launching an Extension Host:

npm run compile
node test-runtime.js

3. Run & Debug the Extension in VS Code

1. Open this repository folder in VS Code (code .).
2. Ensure the project is compiled (npm run compile or keep npm run watch running in a terminal).
3. Launch the Extension Development Host:
   • Press F5 (or go to Run and Debug Ctrl+Shift+D / Cmd+Shift+D -> select Run Extension).
   • .vscode/launch.json is pre-configured to launch this extension and automatically disable the original netcomlabs-ai.copilot-cost-token-tracker extension in the debug window so they do not conflict.
4. In the newly opened [Extension Development Host] window, open the dashboard using any entry point:
   • Click $(zap) Token Cost in the bottom Status Bar
   • Click the Token Cost Tracker icon in the Activity Bar
   • Open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P) and run Token Cost Tracker for Copilot & Ollama Cloud: Open Dashboard

Debugging Tips

• Extension Host (Backend TypeScript): Set breakpoints directly in src/*.ts files (e.g., src/extension.ts, src/dashboardPanel.ts, src/sessionParser.ts, src/sessionDiscovery.ts). Output and console logs appear in the Debug Console of your main VS Code window.
• Dashboard Webview (Frontend HTML/JS): In the [Extension Development Host] window, open the Command Palette and run Developer: Open Webview Developer Tools to inspect the DOM, styles, and console logs for media/dashboard.html.
• Reloading Changes:
  • For changes in media/dashboard.html or media/pricing.json: close and reopen the dashboard panel (or run Developer: Reload Window in the Extension Development Host).
  • For changes in src/*.ts: ensure npm run compile (or npm run watch) has completed, then press Ctrl+R / Cmd+R (or click the Restart button in the debug toolbar) to reload the Extension Development Host.

4. Package & Install Locally (.vsix)

To build a standalone .vsix package and install it in your local VS Code instance:

npm run package

This compiles the TypeScript code and outputs copilot-ollama-cost-tracker-<version>.vsix in the project root. You can install it via the UI (Extensions -> ... -> Install from VSIX...) or from the CLI:

code --install-extension copilot-ollama-cost-tracker-1.1.5.vsix

Supported Ollama Cloud Models & Tags

Built for developers using Ollama Cloud (ollama.com/pricing, Ollama Pro $20/mo with $60 AI Credits, Ollama Max $100/mo with $300 AI Credits) alongside GitHub Copilot Chat (including BYOK / Custom Endpoint / Ollama provider integrations in VS Code):

• Moonshot AI (Kimi): kimi-k2.5:cloud, kimi-k2-thinking:cloud, kimi-k3:cloud, kimi-k2.7-code
• Zhipu AI (GLM): glm-5.1:cloud, glm-5:cloud, glm-4.7:cloud, glm-4.6:cloud
• DeepSeek: deepseek-v3.2:cloud, deepseek-v3.1:671b:cloud
• Alibaba (Qwen): qwen3.5:397b:cloud, qwen3.5:cloud, qwen3-coder-next:cloud, qwen3-coder:480b:cloud, qwen3-next:80b:cloud, qwen3-vl:235b:cloud
• MiniMax: minimax-m2.7:cloud, minimax-m2.5:cloud, minimax-m2.1:cloud, minimax-m2:cloud
• Google / Mistral / NVIDIA / OpenAI OSS: gemma4:31b:cloud, devstral-2:123b:cloud, devstral-small-2:24b:cloud, ministral-3:14b:cloud, nemotron-3-super:cloud, nemotron-3-nano:30b:cloud, gpt-oss:120b:cloud, gpt-oss:20b:cloud

Tags: ollama, ollama-cloud, ollama cloud, ollama pricing, ollama cost, ollama tokens, ollama credits, ollama usage, ollama pro, ollama max, github-copilot, copilot-chat, byok, token-tracker, cost-tracker, ai-credits, llm-cost

Credits & Attribution

Forked from https://github.com/assaelaz/copilot-cost-token-tracker.

License

MIT