> For the complete documentation index, see [llms.txt](https://docs.robinmesh.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.robinmesh.com/quickstart.md).

# Quickstart

There are three ways into RobinMesh: as a user, as a developer, or as a provider. All three work today, and none of them asks for anything more than an Ethereum wallet.

***

## Users: run your first inference

**Prerequisites:** An Ethereum wallet (MetaMask, Rabby, and Robinhood Wallet all work) holding a small amount of USDG. You will not need ETH at any point; RobinMesh sponsors gas through ERC-4337.

**Takes:** Less than 5 minutes.

**1. Connect a wallet**

Open [robinmesh.com/app](https://robinmesh.com/app) and hit "Connect wallet." That is the entire onboarding. No email address, no registration form.

**2. Fund some credits**

Inference is priced in credits, where one credit is worth $0.01 USDG. A starting balance can be as small as $1.

Choose "Add credits," pick an amount, and confirm the USDG transaction in your wallet. The credits show up right away.

**3. Pick a model and chat**

Select any model from the list. A good default is **Qwen3 8B** (Standard tier, 8 credits per request): quick, capable, and hosted by enough workers that it is always reachable.

Send a message. Encryption happens in your browser before the prompt goes anywhere, then the job lands on an available worker, inference runs, and tokens stream back to your screen.

Done. No log was written anywhere, and the payment for the job settled on Robinhood Chain.

***

## Developers: hit the API

**Prerequisites:** A wallet holding USDG, plus an API key generated from the dashboard.

**Takes:** Less than 10 minutes.

**1. Generate an API key**

Sign in at [robinmesh.com/app](https://robinmesh.com/app), open Settings, and create a key. Keys look like `rmesh_live_...` and draw on the credit balance of the wallet that created them.

**2. Send a request**

The API speaks the OpenAI wire format, so existing OpenAI SDK code needs only a new base URL:

```python
from openai import OpenAI

client = OpenAI(
    base_url="https://api.robinmesh.com/v1",
    api_key="rmesh_live_your_key_here"
)

response = client.chat.completions.create(
    model="llama-3.3-70b",
    messages=[{"role": "user", "content": "Explain optimistic rollups in plain terms."}],
    stream=True
)

for chunk in response:
    print(chunk.choices[0].delta.content or "", end="", flush=True)
```

```typescript
import OpenAI from "openai"

const client = new OpenAI({
    baseURL: "https://api.robinmesh.com/v1",
    apiKey: "rmesh_live_your_key_here"
})

const stream = await client.chat.completions.create({
    model: "qwen3-8b",
    messages: [{ role: "user", content: "What is an ERC-4337 smart account?" }],
    stream: true
})

for await (const chunk of stream) {
    process.stdout.write(chunk.choices[0]?.delta?.content ?? "")
}
```

**3. Verify the receipt on-chain**

Each API call produces a transaction on Robinhood Chain. Look for the `x-robinmesh-job-id` and `x-robinmesh-tx-hash` response headers; the latter is the transaction hash, which you can look up on Blockscout at `robinhoodchain.blockscout.com` to confirm the payment yourself.

***

## Providers: put a GPU to work

**Prerequisites:** A browser with WebGPU support (Chrome 113 or newer) and an Ethereum wallet. Browser workers need no USDG to begin.

**Takes:** Less than 5 minutes.

**Browser worker (nothing to install)**

1. Open [robinmesh.com/app](https://robinmesh.com/app) and choose "Earn."
2. Connect your wallet.
3. The page checks your browser for WebGPU support and reads the available VRAM.
4. Pick a model your hardware can run.
5. Press "Start Earning."

Within seconds, jobs start arriving. Each completed job pays USDG straight to your wallet, and you are free to close the tab whenever you like.

**Native worker (better payouts)**

The \[Native Worker guide]\(

) walks through installation and configuration. Running natively pays more per job, opens up larger models, and makes you eligible for the staking multiplier, which moves your payout share from 75% to 85%.

***

## Launch model lineup

| Model         | Tier     | Cost per request   | VRAM required |
| ------------- | -------- | ------------------ | ------------- |
| Llama 3.3 70B | Max      | 40 credits ($0.40) | 48GB+         |
| DeepSeek R1   | Max      | 40 credits ($0.40) | 48GB+         |
| Qwen3 8B      | Standard | 8 credits ($0.08)  | 8GB+          |
| Mistral 7B    | Standard | 8 credits ($0.08)  | 8GB+          |
| Llama 3.2 3B  | Lite     | 2 credits ($0.02)  | 4GB+          |
| Qwen3 1.7B    | Lite     | 2 credits ($0.02)  | 3GB+          |

Once a completion runs past roughly 500 output tokens, billing shifts from a flat per-request price to a per-1,000-output-token rate.

***

## Keep going

* \[Core Concepts]\() explains the network from job submission to settlement
* \[API Reference]\() documents every endpoint
* \[Providers: Browser Worker]\() covers browser worker setup in depth
* \[Providers: Native Worker]\() is the complete native installation guide
* \[Staking]\() shows how to unlock the 85% payout multiplier


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.robinmesh.com/quickstart.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.