# OpenCode

Connect → OpenCode

# Use AxForge with OpenCode

    OpenCode is an open-source coding agent for the
    terminal. It takes any OpenAI-compatible service as a provider in its config file.

    **Before you start**

        - An AxForge API key. Get one (a minute).

        - Node.js 18 or newer. This guide was tried with OpenCode 1.19.3 on 2026-10-17.

## Set it up

### Install OpenCode

```
npm install -g opencode-ai
```

### Add AxForge as a provider

Create `~/.config/opencode/opencode.json` (on Windows
        `%USERPROFILE%\.config\opencode\opencode.json`):

```
{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "axforge": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "AxForge",
      "options": {
        "baseURL": "https://api.axforge.ai/v1",
        "apiKey": "{env:AXFORGE_API_KEY}"
      },
      "models": {
        "qwen3.8-27b-nvfp4": { "name": "Qwen3.8 27B", "limit": { "context": 262144, "output": 16384 } },
        "glm-5.3-flash": { "name": "GLM-5.3-Flash", "limit": { "context": 262144, "output": 16384 } },
        "mimo-v2.6-flash": { "name": "MiMo-V2.6-Flash", "limit": { "context": 262144, "output": 16384 } }
      }
    }
  },
  "model": "axforge/qwen3.8-27b-nvfp4"
}
```

`{env:AXFORGE_API_KEY}` reads the key from the environment, so it never lands in the file. The
        last line makes Qwen3.8 27B the model OpenCode starts with. Each model's `limit` tells OpenCode
        that it reads 262,144 tokens at once, so it shortens a long session before it reaches that.

### Put your key in the environment

```
export AXFORGE_API_KEY="orx_live_your_key_here"
```

```
$env:AXFORGE_API_KEY = "orx_live_your_key_here"
```

It lasts until you close this terminal. Start OpenCode from the same one.

### Start it in your project

```
cd my-project
opencode
```

Under the message box it says **Qwen3.8 27B AxForge**.

### Give it a task

Ask in plain words. It finds the file, edits it and shows the result.

## The settings

        | Setting | Value |  |

        | npm | @ai-sdk/openai-compatible |  |

        | options.baseURL | https://api.axforge.ai/v1 |  |

        | options.apiKey | {env:AXFORGE_API_KEY} |  |

        | models | qwen3.8-27b-nvfp4, glm-5.3-flash, mimo-v2.6-flash |  |

        | limit (each model) | context 262144, output 16384 |  |

        | model | axforge/qwen3.8-27b-nvfp4 |  |

Which model for what: `qwen3.8-27b-nvfp4` is the
    everyday default and reads images, `glm-5.3-flash` thinks before it answers,
    `mimo-v2.6-flash` is the fastest. To start on another one, change the last line of the file, for
    example to `axforge/glm-5.3-flash`.

## Check that it works

    Under the message box it says **Qwen3.8 27B AxForge**, and a task gets done. The calls show in
    the console under API Keys → Recent calls within a minute.

## When it does not work

      | What you see | What it means | Fix |  |

        | "Invalid API key. Create one at https://console.axforge.ai/keys." | the key is not in this terminal, or it was refused | set it as in step 3, in the same terminal; or make a new key |  |

        | Another model or provider under the message box | the config file was not found | check the path and the file name, `opencode.json` |  |

    Test a call says in one click whether a problem is your key,
    your balance or us.

      ← Kilo Code
      n8n →



Source: https://dev.axforge.ai/docs/connect/opencode/