Goose | Vtrix API Docs

Overview

Goose can connect to OpenAI-compatible endpoints. Configure Goose’s OpenAI provider with the Vtrix host, chat completions path, API key, and model ID.

Prerequisites

Goose CLI
A Vtrix API key, created in Settings

Optional: use Vtrix Doctor

You can also use Vtrix Doctor to check the local Base URL, API key, and model settings before editing files. Supported tools, including OpenClaw and Hermes, can be configured after confirmation.

Configuration Steps

Install Goose

• Using the official install script

curl -fsSL https://github.com/aaif-goose/goose/releases/download/stable/download_cli.sh | bash

After installation, run:

goose --version

Configure Vtrix API

Goose’s OpenAI provider uses OPENAI_HOST and OPENAI_BASE_PATH separately. Add the following to ~/.bashrc or ~/.zshrc:

export GOOSE_PROVIDER="openai"
export GOOSE_MODEL="gpt-5.5"
export OPENAI_API_KEY="YOUR_VTRIX_API_KEY"
export OPENAI_HOST="https://cloud.vtrix.ai"
export OPENAI_BASE_PATH="llm/chat/completions"

After updating the file, restart your terminal or run:

source ~/.zshrc

Base URL

The Vtrix OpenAI-compatible base is https://cloud.vtrix.ai/llm. For Goose, set OPENAI_HOST to https://cloud.vtrix.ai and OPENAI_BASE_PATH to llm/chat/completions, which resolves to https://cloud.vtrix.ai/llm/chat/completions.

Choose a Vtrix model

Set GOOSE_MODEL to a Vtrix model ID.

Type	Example model	Notes
High performance	`gpt-5.5`	Strong general coding and reasoning
Standard	`gpt-5.4`	Balanced daily development
Qwen	`qwen3.7-max`	Qwen route through the same Vtrix endpoint

To switch models:

export GOOSE_MODEL="qwen3.7-max"

Tip

View current model IDs on the Models page. Copy the model ID exactly as shown there.

Start a Goose session

Go to your project directory and start Goose:

cd /path/to/your/project
goose session

Ask a simple question, for example:

Summarize this project in one sentence.

If Goose responds normally, requests are reaching Vtrix through the configured endpoint. You can also check usage or billing in Vtrix Settings.

Troubleshooting

Error: 401 Unauthorized

Cause: The Vtrix API key is missing, incorrect, or not loaded in the terminal where Goose started.

Fix:

echo $OPENAI_API_KEY

If the value is empty, check your shell config and restart the terminal.

Requests go to the wrong endpoint

Cause: OPENAI_HOST or OPENAI_BASE_PATH is missing or still points to another provider.

Fix:

env | grep -E 'GOOSE_|OPENAI_'

Confirm that OPENAI_HOST is https://cloud.vtrix.ai and OPENAI_BASE_PATH is llm/chat/completions.

Error: model not found

Cause: GOOSE_MODEL does not match a Vtrix model ID.

Fix: Open the Models page, copy the correct model ID, and update GOOSE_MODEL.

Goose still uses an old provider

Cause: A previous Goose configuration may still be selected, or the current shell did not load the new environment variables.

Fix: Restart the terminal and run:

goose configure

Select the OpenAI provider if Goose asks which provider to use.