> ## Documentation Index
> Fetch the complete documentation index at: https://docs.trynito.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Your First Fusion

> Send one prompt to two or three models and get one synthesized answer.

<Note>
  Fusion needs a **paid plan**. See [What Paid Unlocks](/plans/what-paid-unlocks).
</Note>

`fusion` sends one prompt to two or three models at once, then a judge model synthesizes their answers into one.

## Run It

Name two or three models, then put your prompt after `--`:

<Tabs>
  <Tab title="Claude Code">
    ```text theme={"theme":{"light":"github-light","dark":"github-dark"},"languages":{"custom":["typescript","python","curl"]}}
    /nito:fusion qwen/qwen3-coder z-ai/glm-5.2 -- Should a high-contention inventory counter use optimistic or pessimistic locking? Answer in 3 sentences.
    ```
  </Tab>

  <Tab title="Codex">
    ```text theme={"theme":{"light":"github-light","dark":"github-dark"},"languages":{"custom":["typescript","python","curl"]}}
    $nito:fusion qwen/qwen3-coder z-ai/glm-5.2 -- Should a high-contention inventory counter use optimistic or pessimistic locking? Answer in 3 sentences.
    ```
  </Tab>
</Tabs>

Fewer than two models or more than three is refused, and they have to be distinct.

## Reading the Result

The result has two parts: each model's own answer, listed separately, then a synthesis that combines them. The judge resolves disagreements where it can and flags material ones where it cannot.

Reading the participants alongside the synthesis is the reason to run Fusion. Three models converging is a different signal from three models splitting, and you can only see that if you read them.

## Pick Models That Differ

Models from one family tend to reach the same answer the same way, so a panel of near-relatives agrees with itself and tells you nothing. Name models from different families to get a real second opinion.

## Privacy Levels Have to Be Compatible

Anonymous and Private models can mix in one Fusion. A Confidential model has to stay in a Confidential-only Fusion, so a synthesis never quietly downgrades a Confidential participant.

## Next

<CardGroup cols={2}>
  <Card title="Fusion (full reference)" icon="layer-group" href="/commands/fusion">
    Every option, including judge selection and attachments.
  </Card>

  <Card title="Choosing a Model" icon="list-check" href="/commands/choosing-a-model">
    How to pick strong, complementary models for a panel.
  </Card>
</CardGroup>