tq-serve 0.4.25

TurboQuant model server manager — auto-configured llama-server with KV cache compression

tq — TurboQuant Model Server Manager

Auto-configured llama-server with KV cache compression.

Install

# macOS (Apple Silicon) — one-liner
curl -fsSL https://raw.githubusercontent.com/xt8086/tq/main/install.sh | bash

Note: tq-serve is a pure Python package — it does NOT include the llama-server binary. The binary comes from TheTom/llama-cpp-turboquant, a fork of llama.cpp with TurboQuant KV cache compression built in. tq install (run automatically by the installer) downloads the correct binary for macOS Metal.

Requires macOS Apple Silicon.

Quick Start

tq doctor                        # Verify setup
tq list                          # List local GGUF models
tq search "qwen2.5 coder 7b"    # Search HuggingFace (numbered, pick # to download)
tq serve 1                       # Launch with auto-configured TurboQuant
tq serve 1 --lan                 # Allow other devices on WiFi to connect
tq chat                          # Interactive coding agent

How It Works

tq serve automatically:

1. Detects your hardware (GPU, RAM)
2. Parses model metadata (quant type, layers, context length)
3. Calculates optimal TurboQuant cache settings
4. Launches llama-server with the right flags

Example: A Q4_K_M model on Apple M1 with 8GB RAM gets:

• ctk=q8_0 (protect K cache)
• ctv=turbo4 (compress V cache 3.8x)
• Context capped to safe memory limit
• Idle auto-stop after 5 min

Commands

Command	Description
tq list	List local GGUF models
tq search <query>	Search HuggingFace for GGUF models (numbered, with download prompt)
tq download <model>	Download a model from HuggingFace
tq remove <model>	Remove a downloaded model
tq serve 1	Launch with auto TQ config
tq serve 1 --lan	Allow access from other devices on WiFi
tq serve 1 --dry-run	Show command without running
tq status	Check if server is running
tq stop	Stop the server
tq logs	View server logs
tq validate <model>	Pre-flight check
tq install	Download TurboQuant+ binary (from TheTom/llama-cpp-turboquant)
tq doctor	Verify setup
tq config show	Show/edit configuration
tq chat	Interactive coding agent (local AI)

API

The server exposes an OpenAI-compatible API:

POST http://127.0.0.1:8080/v1/chat/completions

No auth needed. Works with any OpenAI client:

from openai import OpenAI
client = OpenAI(base_url="http://127.0.0.1:8080/v1", api_key="not-needed")
response = client.chat.completions.create(
    model="your-model.gguf",
    messages=[{"role": "user", "content": "Hello"}]
)

Configuration

Config stored at ~/.tq/config.toml:

tq config show              # Show all settings
tq config set port 9090     # Change port
tq config set idle_timeout 600  # 10 min idle timeout (0 to disable)

TurboQuant Cache Types

Type	Bits	Compression	Use Case
f16	16	1x	No compression (baseline)
q8_0	8	2x	Safe for K cache
turbo4	4.25	3.8x	Best quality/compression for V
turbo3	3.25	4.9x	Aggressive, for large models

Requirements

• Python 3.10+
• macOS (Apple Silicon)
• ~2GB free RAM minimum (depends on model)