ChatGPT is mostly used as a general assistant: answering a customer, rewriting a text, going through an email, explaining a document. Open language models running on your own server or in your own cloud cover the same set of tasks: correspondence and documents never leave your perimeter, you are not billed per request, and the model can be fine-tuned on your own data and internal rules. What you pay for instead is hardware and setup work: a server with a GPU, installation, updates and someone who owns it. On long multi-step tasks an open model usually answers less reliably than a flagship cloud service, so for customer replies and document work it almost always needs fine-tuning and guardrails. The choice comes down to three things: how well the model handles the languages you need, what its license allows, and which model size your hardware can carry.
A family of language models with strong Russian language support, from small versions for a laptop to a flagship on par with commercial APIs.
Sber open models with strong Russian language support and local context, from 10B-A1.8B to 702B, all MIT. GigaChat3.1-Audio handles recordings up to two hours; GFusion is a fast diffusion text version.
T-Bank models fine-tuned from Qwen for Russian: they write and reason in Russian noticeably better than the original. T-Lite is 8B, T-Pro 32B on one GPU; T-Search is a multi-step search agent in Russian and English.
The models that started mass open source in AI. A huge ecosystem of fine-tuned versions and tools.
A reasoning model that thinks step by step before answering. Strong at calculations, logic and code; compact distilled versions are available.
The licenses differ: GigaChat is MIT, T-Pro and DeepSeek-R1 are marked as allowing commercial use, while parts of the Qwen and Llama lines run under the developer's own terms. Have the specific version and its license reviewed by a lawyer before a commercial rollout.
An open model can run on your own server: data stays in-house, there is no per-request fee, and the model can be fine-tuned on your documents.
