JoinGrove

Models Overview

Source: docs/models/overview.md
Generated: 01/09/2026

JoinGrove supports local Ollama models, Copilot BYOK, and optional cloud platform models — with PPP pricing for teams.

Key Concepts

  • What: Model strategy covers offline Qwen/Ollama tiers, signed-in Copilot cloud models, and adaptive RAM-based defaults.
  • Why: Individual developers should code for free offline; teams can opt into cloud sync and shared sessions at PPP rates.
  • Related: Ollama integration (../integrations/ollama.md), comparisons (comparisons.md), pricing (pricing.md).

API / Reference

Model path Cost Network Best for
Ollama local (Qwen2.5-Coder) Free Loopback only Individuals, offline labs
Copilot BYOK + Ollama Free (local) / Copilot sub (cloud) Configurable Hybrid teams
Platform cloud models Team/Campus tier Opt-in egress Studio cloud projects
Adaptive lite/standard Auto by RAM Local 8 GB vs 16 GB+ laptops

Key commands

Command ID
Set up Ollama joingrove.setupOfflineAi
Show pricing (PPP) joingrove.showPricing
Toggle zero egress joingrove.toggleZeroEgressMode
Offline providers table joingrove.openDocsReference → offline-providers

Setup & Installation

  1. Offline path → npm run setup-offline-ai (see Ollama integration doc).
  2. Cloud path → sign in to Copilot and/or joingrove.platformSignIn for Studio cloud.
  3. Review PPP pricing → joingrove.showPricing.
  4. Open reference table Offline AI providers via joingrove.openDocsReference.

Code Examples

Offline setup:

ollama pull qwen2.5-coder:7b
npm run setup-offline-ai

Check adaptive performance profile (auto on startup):

npm run verify-global-south

Next: Read routing.md for how JoinGrove picks models offline vs cloud.

How It Connects