> ## Documentation Index
> Fetch the complete documentation index at: https://docs.gcore.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Limits and quotas

> Gclaw limits and quotas cover model context and output token limits, rate limits, and beta instance limits.

Gclaw uses the DeepSeek V4 Flash model for inference, running on Gcore H200 GPUs.

## Model limits

| Limit | Value |
| - | - |
| Context window | 200,000 tokens |
| Maximum output | 32,000 tokens |
| Reasoning | Enabled |

## Rate limits

Rate limits are managed at the Gcore infrastructure level and depend on the account tier. Specific values are not published and may vary. When a rate limit is exceeded, the API returns an error response — reduce request frequency and retry. If errors persist, contact [Gcore support](mailto:support@gcore.com).

## Instance limits

During the beta period, each account is limited to one Gclaw instance.