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

# What is Mako?

> A 32-billion parameter AI agent with native blockchain access, real-time web intelligence, and autonomous tool orchestration — built on Base.

<Frame>
  <img src="https://mintcdn.com/deepmako/TIWp1B5WZuI5kEnK/images/shark.png?fit=max&auto=format&n=TIWp1B5WZuI5kEnK&q=85&s=b1c8f4fd8631660c85cba9c9ca03e2f9" alt="Mako" style={{ maxWidth: '280px', margin: '0 auto', display: 'block' }} width="1024" height="1024" data-path="images/shark.png" />
</Frame>

Autonomous on-chain tools. OpenAI-compatible API. One call, full answer.

*Built for the deep.*

<CardGroup cols={2}>
  <Card title="Quickstart" icon="bolt" href="/quickstart">
    Make your first API call in 30 seconds.
  </Card>

  <Card title="Base Intelligence" icon="globe" href="/base">
    Curated knowledge of the Base ecosystem.
  </Card>

  <Card title="Use Cases" icon="lightbulb" href="/use-cases">
    Token research, wallet intel, due diligence, and more.
  </Card>

  <Card title="API Reference" icon="code" href="/api-reference/chat-completions">
    Interactive playground to test every endpoint.
  </Card>
</CardGroup>

***

## Autonomous tool orchestration

Mako resolves compound queries by chaining tools in sequence. The gateway handles all tool execution server-side — up to 8 rounds per request — and returns a single final response.

```
"what is happening with Virtuals launches on Base?"
```

Execution trace:

1. `knowledge_search("virtuals")` → Virtuals Protocol context and contract details
2. `web_search("virtuals protocol base launches")` → recent launch activity
3. Synthesis → grounded answer combining curated knowledge and live data

No client-side tool handling required. The gateway orchestrates the full chain and returns the synthesized answer.

## Tool suite

16 read-only tools across four categories:

<CardGroup cols={2}>
  <Card title="On-Chain (9)" icon="link-simple">
    ETH balances, ERC-20 balances and metadata, gas prices, block data, transaction counts, contract detection, ENS resolution, transaction details
  </Card>

  <Card title="Web & Social (4)" icon="globe">
    Web search with multi-provider fallback, full page extraction, tweet reading, music search
  </Card>

  <Card title="Market Data (1)" icon="chart-bar">
    Live crypto prices via CoinGecko
  </Card>

  <Card title="Knowledge & Utility (2)" icon="book">
    RAG over 15+ Base ecosystem projects, tool discovery
  </Card>
</CardGroup>

All tools are read-only. Mako cannot broadcast transactions or modify on-chain state.

## Model naming

Use aliases in your requests. Canonical IDs are returned in responses.

| Alias       | Canonical ID         | Description                                                      |
| ----------- | -------------------- | ---------------------------------------------------------------- |
| `conductor` | `mako-32b-conductor` | Primary production model. Tool orchestration, complex reasoning. |
| `operator`  | `mako-8b-operator`   | Compact variant. Streaming, low-latency.                         |

Both aliases and canonical IDs are accepted in the `model` field. If `model` is omitted, the gateway defaults to `conductor`.

## OpenAI-compatible API

The API follows the OpenAI Chat Completions specification. Any client or SDK that supports OpenAI works with Mako by changing the base URL.

<Tabs>
  <Tab title="Python">
    ```python theme={null}
    from openai import OpenAI

    client = OpenAI(
        base_url="https://gateway.deepmako.com/v1",
        api_key="not-needed",
        default_headers={"x-wallet-address": "0xYourWalletAddress"}
    )

    r = client.chat.completions.create(
        model="conductor",
        messages=[{"role": "user", "content": "what is happening on Base today?"}]
    )
    print(r.choices[0].message.content)
    ```
  </Tab>

  <Tab title="JavaScript">
    ```javascript theme={null}
    const res = await fetch("https://gateway.deepmako.com/v1/chat/completions", {
      method: "POST",
      headers: {
        "Content-Type": "application/json",
        "x-wallet-address": "0xYourWalletAddress"
      },
      body: JSON.stringify({
        model: "conductor",
        messages: [{ role: "user", content: "what is happening on Base today?" }]
      })
    });
    const data = await res.json();
    console.log(data.choices[0].message.content);
    ```
  </Tab>

  <Tab title="cURL">
    ```bash theme={null}
    curl -X POST https://gateway.deepmako.com/v1/chat/completions \
      -H "Content-Type: application/json" \
      -H "x-wallet-address: 0xYourWalletAddress" \
      -d '{
        "model": "conductor",
        "messages": [{"role": "user", "content": "what is happening on Base today?"}]
      }'
    ```
  </Tab>
</Tabs>

***

## Open weights

Mako's model weights are available on Hugging Face under permissive licenses:

* **[Mako-32B Conductor](https://huggingface.co/deepmako/Mako-32B-Conductor)** — 32 billion parameters. Primary production model, optimized for tool orchestration and multi-step reasoning.
* **[Mako-8B Operator](https://huggingface.co/deepmako/Mako-8B-Operator)** — 8 billion parameters. Compact variant for streaming and low-latency deployment.

Additional components of the Mako stack will be open sourced over time.