Skip to content
NoviaHub Docs

Desktop apps

WorkBuddy

Connect NoviaHub models to Tencent WorkBuddy with the Custom option under Settings → Models.

Tencent WorkBuddy is a desktop AI agent workstation. In Tencent’s words, it is designed for all workplace roles: describe your needs in one sentence and it plans and carries out the task like a colleague, delivering verifiable results. It comes with built-in models and also lets you add custom models under Settings → Models. This page connects NoviaHub with the Custom option, over the OpenAI-compatible /chat/completions API.

Prerequisites

Section titled “Prerequisites”
  • A NoviaHub account with an API key. See API keys.
  • Funds in your account. See Wallet and top-ups.
  • WorkBuddy installed and signed in to a WorkBuddy account.

Install WorkBuddy

Section titled “Install WorkBuddy”

WorkBuddy is available for macOS (Apple Silicon and Intel) and Windows (x64 and ARM64). See the official Mac installation guide and Windows installation guide.

On macOS, for example:

  1. Download and open the .dmg, then drag the Tencent WorkBuddy icon into the Applications folder.
  2. If macOS says “Tencent WorkBuddy cannot be opened because it is from an unidentified developer” on first launch, go to System Settings → Privacy & Security, scroll down, click Open Anyway and confirm with Open.
  3. After it starts, click Sign In and follow the prompts.

Configure NoviaHub

Section titled “Configure NoviaHub”

  1. Open WorkBuddy’s Settings → Models page and click Add Model in the Custom Models section.

  2. In the Provider (OpenAI-compatible API only) dropdown, choose Custom.

  3. Fill in:

    Field Value
    Endpoint https://noviahub.com/v1/chat/completions
    API Key Your NoviaHub key (starts with sk-)
    Model Name A model ID on NoviaHub, for example deepseek-v4-flash

  4. (Optional) Expand Advanced Settings and tick what the model supports:

    • Tool Calling: the app describes it as “Allow the model to call tools and functions exposed by the agent runtime.”
    • Image Input: tick it only if the model’s Input types include images.
    • Custom Protocol: leave it unticked; see the note below.

  5. Click Test connection. When it shows “Connection succeeded”, click Save. The configuration is stored on this computer.

Verify

Section titled “Verify”

Test connection in the previous step sends one real, very short request (the code caps it at 1 output token), so it costs a tiny amount on NoviaHub. You can also verify in a chat:

  1. In the model selector in the chat window, find the custom model group and pick the model you added.

  2. Send something simple, such as “Introduce yourself in one sentence.” A reply means the setup works.

  3. Check the call and its cost under Usage Logs in the NoviaHub console.

Switch and add models

Section titled “Switch and add models”
  • Switch models in the model selector in the chat window.
  • To use another model, go back to Settings → Models and add another custom model with the same endpoint and API key and a different model name.
  • This setup uses the Chat Completions API, so pick models that carry the Chat label under Models & Pricing. Every model on the site carried it on 2026-09-28.

Troubleshooting

Section titled “Troubleshooting”

Test connection says “The API key is invalid or does not have access to this model.” WorkBuddy shows this for any 401 or 403 response. From NoviaHub, 401 means the key is wrong, or it is disabled, expired or out of quota; see The four statuses. 403 means the account balance has run out (insufficient_user_quota; top up first), or the key has a model limit or an IP allowlist; see API keys.

Test connection says “Unable to connect to the model service.” Check that the endpoint is spelled correctly and that this computer can reach https://noviahub.com.

Notes

Section titled “Notes”
  • The key stays on your computer. WorkBuddy’s docs say custom model settings, including the API key, are saved only in the local workbuddy/models.json and are not uploaded to the cloud.
  • NoviaHub charges for the calls. With a custom model, WorkBuddy only relays your input to the model service you configured; the cost is deducted from your NoviaHub balance.
  • Old configurations still work. Custom models set up earlier in ~/.codebuddy/models.json keep working after the upgrade and can be viewed, edited or deleted in the interface.

References

Section titled “References”

Checked on 2026-09-28 and 2026-09-29: