> ## 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. # OpenCode > Use AIHubMix in OpenCode: install, authorize with an API key, select models, configure models manually, and set a custom base URL on macOS, Linux, Windows. OpenCode now supports major model families such as GPT, Claude, DeepSeek, Gemini, and Qwen. For OpenCode usage scenarios, the following models are recommended: * `coding-minimax-m2.1-free` * `coding-glm-4.7-free` * `claude-sonnet-4-5` * `gpt-5` * `gpt-5-codex` This document explains how to install, configure, and launch OpenCode on different operating systems, and how to integrate the AIHubMix API. *** Before installation, ensure that your system meets the following requirements: * **Node.js ≥ 18.0.0** * Optional package managers for your operating system: * macOS / Linux: `brew` * Arch Linux: `paru` * Windows: `choco`, `winget`, or `scoop` *** ## 1. Installing OpenCode Choose any of the following installation methods based on your operating system and tooling preferences. ### 1. Install via Node.js ```bash theme={null} # Quick install using a CURL script curl -fsSL https://opencode.ai/install | bash # Using NPM npm install -g opencode-ai # Using Bun bun install -g opencode-ai # Using PNPM pnpm install -g opencode-ai # Using Yarn yarn global add opencode-ai ``` ### 2. Install via Homebrew (macOS / Linux) ```bash theme={null} brew install sst/tap/opencode ``` ### 3. Install via Paru (Arch Linux) ```bash theme={null} paru -S opencode-bin ``` ### 4. Windows Installation ```bash theme={null} # Chocolatey choco install opencode # WinGet winget install opencode # Scoop scoop bucket add extras scoop install extras/opencode # Or install via NPM (Node.js required) # Note: Administrator privileges are required npm install -g opencode-ai ``` ## 2. Launching and Logging in to OpenCode After installation, follow the steps below to complete authentication and configuration. ### 1. Log in to OpenCode Run the following command in your terminal: ```bash theme={null} opencode auth login ``` ### 2. Select the AI Provider On the login screen, select **AiHubMix** as the model provider. Opencode 1 ### 3. API Key Authorization Enter your AiHubMix API Key on the OpenCode login page to complete authorization. Opencode 2 *** ## 3. Using Models ### 1. Installation Verification Run the following command to verify that OpenCode is installed and configured correctly: ```bash theme={null} opencode ``` Opencode 3 *** ### 2. Select a Model Type `/models`, search for `aihubmix`, and select the desired model. Opencode 4 *** ## 4. Using AiHubMix in the OpenCode Desktop Client 1. Download the OpenCode desktop application: [https://opencode.ai/download](https://opencode.ai/download) . 2. Open OpenCode and follow the steps:`/model` → `Connect provider` → search for `aihubmix` → enter your API Key → `Submit`. Opencode 5 3. In the model selection panel, search for `aihubmix` and select the model you want to use. Opencode 6 *** ## 5. Manually Configuring and Updating Models This example adds **DeepSeek V4 Pro**. ### Step 1: Confirm the model ID Find the target model on the [AIHubMix](https://aihubmix.com/models) models page and copy its ID. You can also query it directly through the API: ```bash theme={null} curl https://aihubmix.com/v1/models \ -H "Authorization: Bearer YOUR_API_KEY" \ | python3 -m json.tool | grep "id" ``` ### Step 2: Open the configuration file Run the command for your operating system in a terminal: **Mac** ```bash theme={null} mkdir -p ~/.config/opencode && open -a TextEdit ~/.config/opencode/opencode.json ``` **Linux** ```bash theme={null} mkdir -p ~/.config/opencode && nano ~/.config/opencode/opencode.json ``` **Windows** (run in PowerShell) ```powershell theme={null} New-Item -ItemType Directory -Force "$env:USERPROFILE\.config\opencode" notepad "$env:USERPROFILE\.config\opencode\opencode.json" ``` Once the file opens, you will see either an empty file or existing JSON content. ### Step 3: Write the configuration **If the file is empty**, paste the following: ```json theme={null} { "$schema": "https://opencode.ai/config.json", "provider": { "aihubmix": { "models": { "deepseek-v4-pro": { "name": "DeepSeek-V4-Pro", "contextLength": 131072, "type": "language" } } } } } ``` **If the file already has content**, append the new model inside `"models"`, separating each model with a comma: ```json theme={null} "models": { "deepseek-v4-pro": { "name": "DeepSeek-V4-Pro", "contextLength": 131072, "type": "language" }, "deepseek-v4-flash-think": { "name": "DeepSeek-V4-Flash-Think", "contextLength": 131072, "type": "language" } } ``` > **Note**: Do not add a comma after the last model, otherwise the file format will be invalid. Save the file: `Command + S` on Mac, `Ctrl + S` on Windows. **Field reference:** | Field | Description | | --------------- | -------------------------------------------------- | | `name` | Model ID | | `contextLength` | Context length (in tokens) | | `type` | Model type; use `"language"` for text conversation | ### Step 4: Restart OpenCode After saving the configuration file, **fully quit OpenCode and restart it** for the new model to load. ### Step 5: Verify Open the model selector (default shortcut `;`), search for `V4` or `DeepSeek`, and confirm that `DeepSeek-V4-Pro` appears in the list. ***

6. Custom Base URL

By default, OpenCode sends AIHubMix requests to `https://aihubmix.com/v1`. If you need requests to go through a self-hosted gateway, an enterprise proxy, or an acceleration node, you can override this in the configuration file via `options.baseURL`, **without modifying the OpenCode source code or reinstalling**. ### Step 1: Open the configuration file Open `~/.config/opencode/opencode.json` as described in section 5 (on Windows, `%USERPROFILE%\.config\opencode\opencode.json`). ### Step 2: Add `options.baseURL` Add `options.baseURL` under the `aihubmix` provider. You only need to write this single field, and OpenCode will automatically deep-merge it with the built-in configuration, so there is no need to rewrite the entire provider: ```json theme={null} { "$schema": "https://opencode.ai/config.json", "provider": { "aihubmix": { "options": { "baseURL": "https://api.inferera.com" } } } } ``` > **Note**: `baseURL` only needs to be specified at the domain level (e.g., `https://api.inferera.com`). Paths such as `/v1` and `/chat/completions` are appended automatically by the SDK. Do not add `/v1` yourself (e.g., `https://api.inferera.com/v1`), otherwise it will be concatenated into `/v1/v1/` and result in a 404. ### Step 3: Restart and verify Fully quit OpenCode and restart it. You can run the following command to confirm that the configuration has taken effect: ```bash theme={null} opencode debug config ``` You should see the address you entered in the output under `provider.aihubmix.options.baseURL`; from then on, every model request will be sent to that address. **Field reference:** | Field | Description | | ----------------- | -------------------------------------------------------------------------------------------------------------------------------------- | | `options.baseURL` | Custom request address; takes priority over the default address. Specify it at the domain level only, paths are appended automatically | *** ## 7. FAQ and Tips * If the API Key fails validation, log back into the AiHubMix Console and verify that the key is valid: [https://console.aihubmix.com/token](https://console.aihubmix.com/token) * View the built-in help documentation: `opencode` → `/help` * OpenCode official documentation: [https://opencode.ai/docs/](https://opencode.ai/docs/) *** Last updated: 2026-07-29