Translated with AI from the Russian original.

At the same time, the company added a Playground environment for testing and comparing AI models prior to integrating them into products

The MWS Cloud team has launched night rates in the MWS GPT Model Hub service. From 1:00 to 7:00, the cost of using available language models will be reduced to 30% of standard prices. According to the company, this is the first such offering on the Russian market.

The reduced rates apply daily, including weekends and holidays. Companies will be able to shift tasks that do not require an immediate response to nighttime: text generation, document analysis, report preparation, review processing, and embedding generation. Transitioning to the new terms will not require changing existing integrations.

In addition, Playground has been introduced to MWS GPT Model Hub—an environment for testing models prior to connecting them to corporate products. With its help, developers can test queries on their own scenarios, compare response quality, and configure parameters taking processing speed and cost into account.

The service provides access to models for working with text, code, images, and audio, as well as for search and RAG scenarios. Connection is established via an OpenAI-compatible API. Requests are processed on MWS Cloud infrastructure, and data remains within the provider’s perimeter. Available models include GLM-5.3, Kimi K2.6, Qwen3.6 35B (A3B), Gemma 4 31B, and gpt-oss-120b.

Now on the front page