---
title: "YandexGPT (language model) (RO)"
source: "https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)"
wiki: "systems-analysis.info/int"
article: "YandexGPT_(language_model)_(RO)"
language: "ro"
categories:
  - "Category:Large language models"
  - "Category:LLM families"
  - "Category:Machine learning"
  - "Category:Romanian"
revision_id: 8524
wiki_created_at: 2026-09-07T01:18:38Z
wiki_modified_at: 2026-09-07T01:18:38Z
downloaded_at: 2026-09-07T23:25:51Z
---

# YandexGPT (language model) (RO)

**YandexGPT (Yet another GPT)** — familie de modele lingvistice mari, dezvoltate de Yandex și prezentate pentru prima dată în mai 2023.<sup>[\[1\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-rbc-may17-1)</sup> Rețelele YandexGPT sunt utilizate în asistentul integrat Alice, în Căutare și în alte servicii, fiind disponibile și printr-un API public al platformei Yandex Cloud.<sup>[\[2\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-cloud-api-2)</sup>

**YaLM‑100B** (2022) — modelul de cercetare anterior cu 100 de miliarde de parametri și cod sursă deschis; acesta a servit drept „dovadă a conceptului", însă YandexGPT a fost creat separat pentru utilizare comercială.<sup>[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-github-yalm-3)</sup>

## История релизов - Istoricul lansărilor

| Dată        | Lansare                   | Caracteristici cheie                                                                                                                                                                |
|-------------|---------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Iun 2022    | YaLM‑100B                 | 100 B parametri, 1,7 TB date; Apache 2.0.<sup>[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-github-yalm-3)</sup>                              |
| 17 mai 2023 | YandexGPT 1.0             | Integrare în „Alice".<sup>[\[1\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-rbc-may17-1)</sup>                                                    |
| 7 sep 2023  | YandexGPT 2               | +67 % calitate conform testelor interne.<sup>[\[4\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-rbc-sep8-4)</sup>                                  |
| 28 mar 2024 | YandexGPT 3 Pro / Lite    | Nouă linie corporativă API.<sup>[\[5\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-rbc-mar28-5)</sup>                                              |
| 24 oct 2024 | YandexGPT 4 Pro / Lite    | Context de 32 000 de token-uri; raționament implicit (chain‑of‑thought).<sup>[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-habr-gpt4-6)</sup> |
| 25 feb 2025 | YandexGPT 5 Pro           | Paritate cu GPT‑4o în 64 % din sarcini.<sup>[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-adindex-5pro-7)</sup>                               |
| 31 mar 2025 | YandexGPT 5 Lite Instruct | Model de 8 miliarde în acces deschis; format Llama.<sup>[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-hf-5lite-8)</sup>                       |

Versiuni principale

## Архитектура и обучение - Arhitectură și antrenare

- Arhitectură de bază: transformer, optimizat pentru limba rusă.
- YandexGPT 5 Lite: compatibil cu Llama; pre-training ≈ 15 trilioane de token-uri, fine-tuning ulterior ≈ 320 de miliarde.<sup>[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-hf-5lite-8)</sup>

### Контекст и лимиты - Context și limite

- Limita arhitecturală a contextului — 32 000 de token-uri (versiunile 4/5).<sup>[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-habr-gpt4-6)</sup>
- API-ul public limitează o singură cerere (prompt + completion) la 7 400 de token-uri.<sup>[\[9\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-langchain-7400-9)</sup>
- Dimensiunea maximă a \*\*răspunsului\*\* — 2 000 de token-uri conform secțiunii „Quotas and limits".<sup>[\[10\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-cloud-limits-10)</sup>

## Текущие модели (июнь 2025) - Modele actuale (iunie 2025)

| Model            | Parametri | Context | Licență                 | Observații                                                                                                                                   |
|------------------|-----------|---------|-------------------------|----------------------------------------------------------------------------------------------------------------------------------------------|
| YandexGPT 5 Pro  | n/d       | 32 000  | proprietară             | Acces prin API și „Alice Pro".<sup>[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-adindex-5pro-7)</sup> |
| YandexGPT 5 Lite | 8 mld     | 32 000  | Yandex GPT‑Lite License | Deschis; compatibil cu Llama.<sup>[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-hf-5lite-8)</sup>      |
| YaLM‑100B        | 100 mld   | 2 048   | Apache 2.0              | Proiectul original.<sup>[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-github-yalm-3)</sup>             |

## Бенчмарки - Benchmark-uri

- Teste interne: 5 Pro a atins paritatea cu GPT‑4o în 64 % din sarcini; depășire față de 4 Pro — 67 %.<sup>[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-adindex-5pro-7)</sup>
- ru‑LLM Arena: YandexGPT menține poziția de lider după scorul ELO în rândul modelelor în limba rusă.<sup>[\[11\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-llmarena-11)</sup>

## Fine‑tuning

Pentru 5 Lite este suportată oficial metoda LoRA; un exemplu de rulare a fost publicat în model card.<sup>[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-hf-5lite-8)</sup>

## API‑режимы - Moduri API

- Sincron — răspunsuri rapide (Lite).
- Asincron — sarcini cu consum ridicat de resurse (Pro).<sup>[\[2\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-cloud-api-2)</sup>

## Мультимодальность - Multimodalitate

Familia YandexGPT rămâne orientată spre text; serviciile multimodale („Neuro", „YandexArt", „Yandex Vision") sunt dezvoltate separat.<sup>[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_note-habr-gpt4-6)</sup>

## Ссылки - Referințe externe

- YandexGPT în Yandex Cloud — pagina serviciului
- Ponderile YandexGPT‑5 Lite pe Hugging Face
- Depozitul YaLM‑100B pe GitHub

## Литература - Bibliografie

- Matkin, N. et al. (2024). *Comparative Analysis of Encoder-Based NER and Large Language Models for Skill Extraction from Russian Job Vacancies*. arXiv:2407.19816.
- Tsanda, A.; Bruches, E. (2024). *Russian-Language Multimodal Dataset for Automatic Summarization of Scientific Papers*. arXiv:2405.07886.
- Goloburda, M. et al. (2025). *Qorǵau: Evaluating LLM Safety in Kazakh-Russian Bilingual Contexts*. arXiv:2502.13640.
- Togmanov, M. et al. (2025). *KazMMLU: Evaluating Language Models on Kazakh, Russian, and Regional Knowledge of Kazakhstan*. arXiv:2502.12829.
- Noels, S. et al. (2025). *What Large Language Models Do Not Talk About: An Empirical Study of Moderation and Censorship Practices*. arXiv:2504.03803.

## Примечания - Note

1.  <span id="cite_note-rbc-may17-1">↑ <sup>[1.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-rbc-may17_1-0)</sup> <sup>[1.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-rbc-may17_1-1)</sup> «“Яндекс” добавил в “Алису” аналог ChatGPT». *РБК*. <a href="https://www.rbc.ru/technology_and_media/17/05/2023/6464aa4e9a79473a1f0464c1" class="external autonumber" rel="nofollow">[1]</a></span>
2.  <span id="cite_note-cloud-api-2">↑ <sup>[2.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-cloud-api_2-0)</sup> <sup>[2.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-cloud-api_2-1)</sup> «Getting started with YandexGPT (Quickstart)». *Yandex Cloud Docs*. <a href="https://cloud.yandex.com/en/docs/yandexgpt/quickstart" class="external autonumber" rel="nofollow">[2]</a></span>
3.  <span id="cite_note-github-yalm-3">↑ <sup>[3.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-github-yalm_3-0)</sup> <sup>[3.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-github-yalm_3-1)</sup> <sup>[3.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-github-yalm_3-2)</sup> «yandex/YaLM‑100B: Pretrained language model with 100B». *GitHub*. <a href="https://github.com/yandex/YaLM-100B" class="external autonumber" rel="nofollow">[3]</a></span>
4.  <span id="cite_note-rbc-sep8-4">[↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-rbc-sep8_4-0) «Как “Яндекс” решил зарабатывать на своём аналоге ChatGPT». *РБК*. <a href="https://www.rbc.ru/technology_and_media/08/09/2023/64f9ff099a79472ad0e24715" class="external autonumber" rel="nofollow">[4]</a></span>
5.  <span id="cite_note-rbc-mar28-5">[↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-rbc-mar28_5-0) «“Яндекс” представил третье поколение нейросетей YandexGPT». *РБК*. <a href="https://www.rbc.ru/rbcfreenews/660563e29a794766b508c1f9" class="external autonumber" rel="nofollow">[5]</a></span>
6.  <span id="cite_note-habr-gpt4-6">↑ <sup>[6.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-habr-gpt4_6-0)</sup> <sup>[6.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-habr-gpt4_6-1)</sup> <sup>[6.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-habr-gpt4_6-2)</sup> «Более мощное семейство моделей YandexGPT 4». *Habr*. <a href="https://habr.com/ru/companies/yandex/articles/852968/" class="external autonumber" rel="nofollow">[6]</a></span>
7.  <span id="cite_note-adindex-5pro-7">↑ <sup>[7.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-adindex-5pro_7-0)</sup> <sup>[7.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-adindex-5pro_7-1)</sup> <sup>[7.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-adindex-5pro_7-2)</sup> «“Яндекс” внедрил YandexGPT 5 Pro в чат с “Алисой Про”». *AdIndex*. <a href="https://adindex.ru/news/digital/2025/02/25/331154.phtml" class="external autonumber" rel="nofollow">[7]</a></span>
8.  <span id="cite_note-hf-5lite-8">↑ <sup>[8.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-hf-5lite_8-0)</sup> <sup>[8.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-hf-5lite_8-1)</sup> <sup>[8.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-hf-5lite_8-2)</sup> <sup>[8.3](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-hf-5lite_8-3)</sup> «yandex/YandexGPT‑5‑Lite‑8B‑pretrain». *Hugging Face*. <a href="https://huggingface.co/yandex/YandexGPT-5-Lite-8B-pretrain" class="external autonumber" rel="nofollow">[8]</a></span>
9.  <span id="cite_note-langchain-7400-9">[↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-langchain-7400_9-0) «ChatYandexGPT API Reference (max_tokens = 7400)». *LangChain Docs*. <a href="https://python.langchain.com/api_reference/community/chat_models/langchain_community.chat_models.yandex.ChatYandexGPT.html" class="external autonumber" rel="nofollow">[9]</a></span>
10. <span id="cite_note-cloud-limits-10">[↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-cloud-limits_10-0) «Yandex Cloud service quotas and limits → Foundation Models». *Yandex Cloud Docs*. <a href="https://cloud.yandex.com/en/docs/overview/concepts/quotas-limits" class="external autonumber" rel="nofollow">[10]</a></span>
11. <span id="cite_note-llmarena-11">[↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(RO)#cite_ref-llmarena_11-0) «llmarena/llmarena — российская краудсорсинговая платформа оценки LLM». *GitHub*. <a href="https://github.com/llmarena/llmarena" class="external autonumber" rel="nofollow">[11]</a></span>