Skip to main content
Glama
Modelglass

Modelglass

Official
by Modelglass
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

modelglass-vscode

What Modelglass is

Modelglass is a pricing-and-capability data layer for AI models — image, video, audio, and LLM — built as a sourced, append-only registry: every price carries a source URL and the date it was verified, and a repricing is a new dated entry, never a silent overwrite. It's served two ways: a free comparison site, and a paid read API + MCP server (https://modelglass-api.vercel.app, POST /mcp) for operators wiring model pricing into their own tooling. Modelglass itself doesn't proxy your traffic or pick a model at runtime — it's the data other people's tools (including both extensions in this repo) read.

Related MCP server: Artificial Analysis MCP Server

What this repo is

An npm workspaces monorepo publishing two independent VS Code extensions, each installable on its own:

Package

What it is

Marketplace

packages/router

Modelglass Cost-Aware Router — routes dev tasks to the cheapest capable LLM (and generates video via Runway, audio via ElevenLabs) directly against your own provider keys, BYOK, no Modelglass proxy in the request path. Also available directly inside Copilot Chat's model picker.

modelglass.cost-aware-router

packages/mcp

Modelglass MCP — registers the live Modelglass MCP server directly in Copilot's agent mode, no manual .vscode/mcp.json editing. Seven read-only tools (models, pricing history, competitors, account/tier info) — Copilot's agent decides when to use them, this extension just makes them discoverable.

modelglass.modelglass-mcp

packages/shared

Internal only, not published. SecretStorage-backed auth (ensureApiKey/peekApiKey) and the Modelglass API base URL, shared by both extensions above.

—

Both extensions use the same Modelglass account/API key — installing one doesn't require separately authenticating the other. See each package's own README.md for full usage docs, including a screenshot of it in action.

Modelglass: Route Task recommendation

Development

npm install                              # installs + links all three workspace packages
npm run typecheck                        # all packages
npm run test                             # all packages (packages without a test script are skipped)
npm run build                            # all packages

# per-package:
npm run build --workspace packages/router
npm run package --workspace packages/mcp   # -> packages/mcp/*.vsix

To run an extension locally: open packages/router or packages/mcp as the VS Code workspace root and press F5 to launch an Extension Development Host with your changes loaded — standard VS Code extension tooling, no custom launch config needed. See CONTRIBUTING.md for the full local-dev + PR flow.

The router/mcp split is tracked as SCO-434 in the main modelglass repo's Linear project.

Contributing

Bug reports, fixes, and new features are welcome — see CONTRIBUTING.md and our Code of Conduct.

To report a security vulnerability, follow SECURITY.md instead of opening a public issue.

Talk to us

Questions about Modelglass itself or either extension — email scott@modelglass.com.au.

Copyright © 2026 Modelglass Pty Ltd. Licensed under the MIT License — see LICENSE.

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityMaintained
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • Index9 MCP Server

    index9-org
    A
    license
    A
    quality
    F
    maintenance
    Provides AI assistants with real-time access to 1000+ AI models including their latest pricing, context windows, capabilities, and specifications. Supports model search, comparison, recommendations, and live testing.
    3
    26 npm
    1
    MIT

  • Artificial Analysis MCP Server

    davidhariri
    A
    license
    A
    quality
    D
    maintenance
    Provides access to real-time LLM pricing, speed metrics, and performance benchmarks for over 300 models from Artificial Analysis. It enables users to list, filter, and compare models based on costs, tokens per second, and intelligence indices.
    2
    28 npm
    9
    MIT

  • atom-mcp-server

    A7OM-AI
    A
    license
    A
    quality
    D
    maintenance
    Global price benchmarking for AI inference across 2,600+ SKUs from 47 vendors. Query live pricing, market indexes, and model specs via 8 tools. Free tier available.
    8
    78 npm
    MIT