Using the Verisoft LLM Gateway with Zoo Code

Step 1 — Get an API key

Contact your direct manager to request an API key. Keys are issued per person.

Step 2 — Store it securely

Save the key somewhere private on your machine — a password manager or an encrypted note, not a plain file on your desktop. Don't commit it to Git, paste it in chat, or share it with a colleague; they should request their own.

Step 3 — Install the extension

Install Zoo Code from the VS Code marketplace (Extensions → search "Zoo Code" → Install).

Any other extension that supports an API key for an LLM provider works too — the settings below are the same, only field names differ.

Step 4 — Configure the provider

Open the extension's Providers settings and fill in three fields:

Click Refresh Models, then pick a model from the dropdown. Done.

Step 5 — Monthly budget

Every key has a monthly spend limit. Once you hit it, requests fail until the next cycle resets.

Check Monthly Usage

Windows:

$r = Invoke-RestMethod -Uri "https://llm.verisoft.io/key/info" `
    -Headers @{ Authorization = "Bearer <YOUR_KEY_HERE>" }
$r.info | Select-Object spend, max_budget, budget_reset_at

Mac:

curl -s https://llm.verisoft.io/key/info \
  -H "Authorization: Bearer <YOUR_KEY_HERE>" | jq '.info | {spend, max_budget, budget_reset_at}'

What model should you choose?

Start with gpt-6-luna , best for day to day use.

If you need more reasoning for complex tasks (like infra tasks), use kimi-k2.6.

If it’s still not enough, you can choose one of the power models.

Prices Table

LiteLLM Model Tiers

base-models

power-models