OnlyTerp/opengrok

Run any model in Grok Bot — one-command setup, model picker UI, evidence-based provider wire maps, and an update-proof doctor. Not farming you, arming you.

What it solves

opengrok is a configuration sidecar for the Grok Bot desktop application. It solves the problem of "harness mismatch," where using foreign AI models (like Claude, Gemini, or DeepSeek) within Grok Bot often results in poor performance, high token usage, or incorrect reasoning behavior because the model is not receiving the specific prompt shapes and parameters it was trained for.

How it works

The project acts as a bridge between Grok Bot and various AI providers. It uses "provider maps"—verified wire-level configurations—to ensure that parameters like "thinking" modes, effort levels, and fast-lane routing are correctly translated from Grok Bot's requests to the specific requirements of the upstream provider (e.g., xAI, Anthropic, Google, Zhipu).

Key components include:

  • LiquidGlass Observatory: An in-app HUD that allows users to swap models mid-chat, monitor live telemetry (latency, throughput), and track context window usage.
  • Wire Probes: Tools that capture and verify actual API requests to ensure that configuration "knobs" actually change model behavior rather than just returning a success code.
  • Binding Consumer: A patch for cloud-hosted Grok Bot instances to allow them to read the custom model bindings.
  • Codex Everywhere Bridge: A local proxy that converts Chat Completions API calls into Responses API calls for specific model compatibility.

Who it’s for

Users of the Grok Bot desktop application who want to use a variety of different LLMs (local or cloud-based) while maintaining optimal performance and native-feeling behavior for each model.

Highlights

  • Mid-chat Model Swapping: Change the "brain" of an agent without restarting the conversation.
  • Evidence-Based Mapping: Every model configuration is verified via wire captures to ensure parameters are actually honored by the provider.
  • Zero-Install Integration: Injected into the app bundle to survive Grok Bot updates.
  • Live Telemetry: Real-time tracking of TTFT (Time to First Token) and token flow per bot.
  • Context Guardian: A visual indicator to prevent clipping during long conversations (up to 256k window).

Related

  • Dispatch
    AI & Frontier Tech Roundup: Grok Bot Surge, Free Model Credits, and New Benchmarks

    This roundup highlights the rapid expansion of AI agents like Grok Bot, a wave of free model credits for GLM‑5.3 and others, and the introduction of the discoverative AI benchmark TRACES.

  • Project
    lidge-jun/opencodex

    A universal provider proxy that lets you use any LLM (Claude, Gemini, Grok, etc.) with Codex, Claude Code, Claude Desktop, and Grok Build.

  • Dispatch
    xAI Grok Release Notes

    xAI has announced Grok, an AI assistant powered by the Grok-1 LLM that features real-time world knowledge via the X platform and a personality modeled after the Hitchhiker's Guide to the Galaxy.

  • Project
    midudev/canirun.ai

    A hardware detection and recommendation tool that tells users which open-weight AI models will run best on their specific machine based on their CPU, RAM, and GPU.

  • Dispatch
    Grok 4.6 Release Notes: Agentic Coding and Frontier Intelligence

    xAI has released Grok 4.6, a model optimized for long-running agents and complex technical work that matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index.