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

# DeepSeek Harness

> Connect DeepSeek Harness to AIHubMix: install the npm release, configure an OpenAI-compatible provider, add auto or specific models, and test an Agent request.

DeepSeek Harness (`dsh`) is DeepSeek's open-source coding agent harness. It provides workspaces, a terminal, file editing, Skills, planning, and a Web UI. Its OpenAI Chat Completions compatibility lets it connect to AIHubMix.

<Info>
  This guide was tested on August 14, 2026 with DeepSeek Harness `0.1.0-rc.6`. Harness is still a developer preview, so later releases may change fields or UI details.
</Info>

<h2 id="summary">
  Quick summary
</h2>

The integration requires four values:

| Setting                      | Value                     |
| ---------------------------- | ------------------------- |
| Provider ID                  | `aihubmix`                |
| Base URL                     | `https://aihubmix.com/v1` |
| API protocol                 | `openai-completions`      |
| API key environment variable | `AIHUBMIX_API_KEY`        |

After configuration, the Models page shows AIHubMix as configured.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/01-models-aihubmix-configured.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=9a11531222d056d05d6ee6030eddb595" alt="AIHubMix configured on the DeepSeek Harness Models page" width="1280" height="720" data-path="public/en/deepseek-harness/01-models-aihubmix-configured.png" />
</Frame>

<h2 id="requirements">
  What do you need?
</h2>

Before you begin, prepare:

* An active [AIHubMix API key](https://console.aihubmix.com/token)
* Node.js `22.19.0` or later, including Node.js 24+
* Network access to the npm registry and `https://aihubmix.com`

Check Node.js and npm:

```bash theme={null}
node --version
npm --version
```

<h2 id="install">
  How do you install DeepSeek Harness?
</h2>

Install the published npm package without cloning the source repository:

```bash theme={null}
npm install -g @deepseek-ai/dsh
```

Verify the installation:

```bash theme={null}
dsh --version
```

If the shell reports `dsh: command not found`, run `npm config get prefix` and make sure its `bin` directory is included in `PATH`.

<h2 id="start-web-ui">
  How do you start the Web UI?
</h2>

Run the following command from the directory you want to use as the workspace:

```bash theme={null}
dsh web --port 3080
```

Open this URL:

```text theme={null}
http://127.0.0.1:3080
```

On first use, select or add a workspace before sending an Agent request.

<h2 id="configure-provider">
  How do you configure the AIHubMix provider?
</h2>

Open **Settings > Models > Add a custom provider**, then enter the following values:

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/03-add-custom-provider.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=e5f609447d3c979a65235ecc1b0382d0" alt="Custom provider form in DeepSeek Harness" width="1280" height="720" data-path="public/en/deepseek-harness/03-add-custom-provider.png" />
</Frame>

| Field        | Recommended value         | Notes                                                                         |
| ------------ | ------------------------- | ----------------------------------------------------------------------------- |
| Provider ID  | `aihubmix`                | Lowercase identifier; use `aihubmix-gateway` if it is already taken           |
| Display name | `AIHubMix`                | Name shown in the model selector                                              |
| Base URL     | `https://aihubmix.com/v1` | Must include `/v1`                                                            |
| API protocol | `openai-completions`      | OpenAI Chat Completions compatible protocol                                   |
| API key      | Your AIHubMix API key     | Harness stores it as a credential and does not show the plaintext value again |

Harness appends `/chat/completions` to the Base URL. The resulting endpoint is:

```text theme={null}
https://aihubmix.com/v1/chat/completions
```

<Warning>
  Do not put the full `/chat/completions` path in the Base URL field.
</Warning>

<h2 id="add-models">
  Add auto or another model
</h2>

After entering the Base URL and API key, select **Fetch available models**. Harness requests `GET https://aihubmix.com/v1/models` and shows the currently available model IDs.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/07-discover-models.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=7fdeeb8c4ba8400910bcabe1e7b50a78" alt="Available AIHubMix models discovered by DeepSeek Harness" width="1280" height="720" data-path="public/en/deepseek-harness/07-discover-models.png" />
</Frame>

For an initial connection test, you can add only `auto` and set its display name to `Auto`. `auto` is AIHubMix's automatic model selection entry point.

To use a specific model, select it from the fetched list or choose **Add model** and enter the exact model ID. The graphical model selector only shows models registered in the current provider catalog.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/04-add-auto-model.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=0c1e93d53cf9840d5c5885dc192b0d80" alt="Adding the AIHubMix auto model manually" width="1280" height="720" data-path="public/en/deepseek-harness/04-add-auto-model.png" />
</Frame>

Select **Create provider** to save the configuration.

<h2 id="auto-vs-specific-models">
  auto vs specific models
</h2>

| Choice            | When to use it                                                        |
| ----------------- | --------------------------------------------------------------------- |
| `auto`            | Initial connection tests, or automatic model selection by AIHubMix    |
| Specific model ID | A fixed model capability, version, or reproducible result is required |

Both choices use the same provider, Base URL, and API key. Switching models does not require creating the provider again.

<h2 id="environment-variables">
  Store the API key in an environment variable
</h2>

For development machines and CI, Harness can reference an environment variable. Edit `$HOME/.dsh/settings.yaml`:

```yaml theme={null}
llm-pi-ai:
  providers:
    aihubmix:
      displayName: AIHubMix
      apiKeyEnv: AIHUBMIX_API_KEY
      api: openai-completions
      baseURL: https://aihubmix.com/v1
      models:
        - id: auto
          name: Auto
          contextWindow: 1000000
          maxTokens: 32768

agent-default-model:
  provider: aihubmix
  model: auto
```

Set the key before starting Harness:

```bash theme={null}
export AIHUBMIX_API_KEY='sk-***'
dsh web --port 3080
```

When the key comes from the launch environment, the settings page displays a read-only message without revealing the value.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/02-aihubmix-provider-details.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=a18af7f6409268331dcb4ddfa7d78568" alt="AIHubMix provider details in DeepSeek Harness" width="1280" height="720" data-path="public/en/deepseek-harness/02-aihubmix-provider-details.png" />
</Frame>

Restart Harness after changing the environment variable.

<h2 id="verify-chat">
  Select a model and test a conversation
</h2>

Close Settings, open the model selector at the lower right of the prompt box, and choose **Auto** or another model you added.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/05-select-auto-model.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=82577ac5295eebc0ef6e4945f75f5ffe" alt="Auto selected in the DeepSeek Harness model selector" width="1280" height="720" data-path="public/en/deepseek-harness/05-select-auto-model.png" />
</Frame>

Send a minimal test message:

```text theme={null}
Reply with exactly: AIHubMix connection successful
```

Receiving the expected response confirms the connection between Harness, the OpenAI Chat Completions compatible endpoint, and AIHubMix.

<Frame>
  <img src="https://mintcdn.com/aihubmix/pPBGbuf_Qo9bMEMd/public/en/deepseek-harness/06-live-chat-success.png?fit=max&auto=format&n=pPBGbuf_Qo9bMEMd&q=85&s=f5fe3902c1c2f57b84b34b739b258394" alt="Successful AIHubMix response in DeepSeek Harness" width="1280" height="720" data-path="public/en/deepseek-harness/06-live-chat-success.png" />
</Frame>

<h2 id="headless-test">
  Run a headless test
</h2>

You can also test without the Web UI:

```bash theme={null}
export AIHUBMIX_API_KEY='sk-***'
dsh --profile headless \
  'Reply with exactly DSH_AIHUBMIX_OK and do not use tools.'
```

A successful run prints `DSH_AIHUBMIX_OK`.

<h2 id="troubleshooting">
  FAQ
</h2>

<h3 id="missing-credential">
  Why does Harness show MISSING\_CREDENTIAL?
</h3>

Harness did not receive `AIHUBMIX_API_KEY` from its launch environment. Set the variable and launch Harness from the same terminal:

```bash theme={null}
test -n "$AIHUBMIX_API_KEY" && echo configured
```

<h3 id="models-401">
  Why does model discovery return 401?
</h3>

Check that the API key is valid and does not contain extra spaces or line breaks. The model discovery request uses Bearer authentication.

<h3 id="only-auto">
  Why is Auto the only model in the selector?
</h3>

The selector shows only models registered in the current provider catalog. Open **Settings > Models > AIHubMix > Edit**, then use **Fetch available models** or add an exact model ID manually.

<h3 id="unknown-model">
  Why does Harness report UNKNOWN\_MODEL?
</h3>

The current provider catalog does not contain that model ID. Model IDs are case-sensitive, so use the exact value returned by model discovery.

<h3 id="connection-timeout">
  Why does the connection time out?
</h3>

Check DNS, your organization network, and local proxy settings. If the browser uses a system proxy but the terminal does not, configure `HTTP_PROXY`, `HTTPS_PROXY`, and `NODE_USE_ENV_PROXY=1` in the terminal that launches `dsh`.

<h2 id="security">
  Security guidance
</h2>

* Do not commit API keys to Git or include them in screenshots and shared scripts
* Prefer Harness credential storage or the `AIHUBMIX_API_KEY` environment variable
* Do not print a complete API key while troubleshooting
* Use the CI platform's secret management feature to inject environment variables

***

Last updated: August 14, 2026