OptiQ Cloud

Give the hard tasks to a cloud model, and keep your Mac doing the work.

Boost runs a task with optiq-boost, the OptiQ Cloud model, in charge. It plans, decides and makes the important edits, and hands the reading, searching and test runs to the model on your Mac. On 30 benchmark tasks, solved tasks went from 4 to 15, for $2.80 in total.

Start with $2 free Read the docs

Boost runs through OptiQ on an Apple Silicon Mac. No card needed to start.

How it works

Turn on Boost before a hard task

Boost is a model you pick, not a mode you get stuck in. Pick it for the task that needs it and go back to your local model for the rest.

Install OptiQ

On an Apple Silicon Mac. Skip this if you already run optiq serve, OptiQ Code or the Lab.

$ pip install -U mlx-optiq

Connect the machine

$ optiq cloud login

Pick Boost

In OptiQ Code, type /boost before the task. The cloud model leads it and calls your local model through its local tool. In any other client pointed at optiq serve, choose the model optiq-boost.

/boost
add retry with backoff to the HTTP client and make the tests pass

See what it cost

The run summary shows what the task cost and what is left on your balance. Type /boost again to go back to your local model.

OptiQ Cloud $0.06 · $1.94 left

Results

What it did on 30 real tasks

The same 30 tasks from DeepSWE-mini and Terminal-Bench mini, the same local Qwen3.6-35B-A3B on a 24 GB Mac mini, run in OptiQ Code with and without Boost.

	Local model alone	With Boost
Tasks solved	4 / 30	15 / 30
DeepSWE-mini target tests	313 / 771	706 / 771
Terminal-Bench tasks solved	4 / 14	8 / 14
Cost for all 30 tasks	free, local	$2.80

One attempt per task. The full write-up, including what the local model did and what it cost per task, is on the blog.

Where it works

It works in the agent you already use

optiq serve and the Lab list optiq-boost next to your local model once you are signed in. Pick it in any client and that conversation runs on OptiQ Cloud. OptiQ Code is the one that also gives the cloud model your local model as a helper.

OptiQ Code

/boost before a task. The cloud model leads and hands bulky work to your local model.

Claude Code

Point ANTHROPIC_BASE_URL at optiq serve and set the model to optiq-boost.

Codex

Chat Completions or Responses, streaming or not. Set the model to optiq-boost.

OpenCode

Add optiq-boost to your optiq serve provider and pick it.

OptiQ Lab

Choose optiq-boost in the chat's model list. Tools and files work as they do locally.

Anything OpenAI-compatible

OpenClaw, Hermes Agent, Mistral Vibe and the rest: the same model name on the same endpoint.

Pricing

What it costs

You pay for the tokens a task uses, from a balance you top up when you want. There is nothing recurring to cancel, and nothing to pay on the days you do not use it.

optiq-boost	Per million tokens
Input	$0.15
Cached input	$0.03
Output	$0.50

Free

$2 of credit

Good for 14 days. On our benchmark run that covered about 35 typical tasks. No card needed.

Credit

$20 at a time

A one-off purchase, not a subscription. Auto top-up is optional, with a monthly cap you set.

Privacy

What leaves your machine

What leaves my machine?

The conversation of a task you run with Boost: your request, the cloud model's replies, and the tool results it asks for, including the parts of files it reads. Tasks on your local model never reach the network. Your repository is not uploaded, indexed or scanned, and the jobs the cloud model hands to your local model stay on your Mac.

Is my code used for training?

No. Prompts and replies are not stored by OptiQ Cloud, and Boost only uses model providers that do not train on requests. What is kept is token counts, timings and cost, so you can see what a task cost.

What if my balance runs out?

The run stops and says so. Nothing is charged past an empty balance except the call already in progress, and your local model keeps working without the cloud.

Can I turn it off?

It is off until you pick it. optiq cloud logout revokes the token rather than just forgetting it.

Try it on your next hard task

$2 of credit to start. One command to connect, and no plugin for whichever agent you already use.

Start with $2 free Read the docs