mlx-optiq

OptiQ Cloud

OptiQ Cloud is an OpenAI-compatible API for three open models chosen for coding agents, paid per token from a dollar balance. You can use it from your coding agent, call the API directly, or use Boost in OptiQ Code.

Base URL https://cloud.mlx-optiq.com/api/v1

Three ways to use it

Models

ModelIdInputCached inputOutputContextTypical speed
DeepSeek V4.1 Flashdeepseek-v4.1-flash$0.10$0.03$0.501M—
GLM 5.3 Flashglm-5.3-flash$0.15$0.03$0.501M—
MiniMax M3.1 Flashminimax-m3.1-flash$0.15$0.02$0.601M—
Qwen3.8 27Bqwen3.8-27b$0.15$0.03$2.00256K—
MiMo V2.6 Flashmimo-v2.6-flash$0.22$0.01$0.65256K—
MiniMax M3minimax-m3$0.50$0.10$2.00512K—
GLM 5.3glm-5.3$1.40$0.26$4.401M—
DeepSeek V4 Prodeepseek-v4-pro$2.20$0.08$6.601M—

Prices are per million tokens, and output includes reasoning. GLM 5.3 and DeepSeek V4 Pro are the larger models, for the hardest tasks, and MiniMax M3 sits in between; the four Flash models and Qwen3.8 27B are fast and cheap. Boost runs on GLM 5.3 Flash unless you pick another model. A long agent run resends its conversation on every call and most of it is served from cache, so the cached price matters most.

How the models are served

All OptiQ Cloud models are served at full precision, and your requests are never used for training.

Connect a machine

Run this once per machine. The browser approves it, and the key is saved in your OptiQ config on that machine.

$ optiq cloud login

optiq cloud status shows the account and balance. optiq cloud logout revokes the key rather than only forgetting it, so a lost laptop cannot keep spending.

What leaves your machine

The conversation of a task on a cloud model: your request, the replies, and the tool results it asks for, including the parts of files it reads. Tasks on your local model never reach the network. Neither do the jobs Boost hands to your local model, only their short answers. Your repository is not uploaded, indexed or scanned.

OptiQ Cloud does not store prompts or replies. It keeps token counts, timings and cost, so the dashboard can show what a task cost.

In this section