--- title: "YandexGPT (language model) (TL)" source: "https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)" wiki: "systems-analysis.info/int" article: "YandexGPT_(language_model)_(TL)" language: "tl" categories: - "Category:Large language models" - "Category:LLM families" - "Category:Machine learning" - "Category:Tagalog" revision_id: 8527 wiki_created_at: 2026-09-07T01:18:40Z wiki_modified_at: 2026-09-07T01:18:40Z downloaded_at: 2026-09-07T23:25:52Z --- # YandexGPT (language model) (TL) **YandexGPT (Yet another GPT)** — pamilya ng malalaking language model na binuo ng Yandex at unang ipinakita noong Mayo 2023.^{[\[1\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-rbc-may17-1)} Ang mga network ng YandexGPT ay ginagamit sa built-in na assistant na Alice, sa Search at iba pang mga serbisyo, at available din sa pamamagitan ng pampublikong API ng Yandex Cloud platform.^{[\[2\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-cloud-api-2)} **YaLM‑100B** (2022) — isang naunang research model na may 100 bilyong parameter at bukas na code; nagsilbi itong "patunay ng konsepto", ngunit ang YandexGPT ay binuo nang hiwalay para sa komersyal na paggamit.^{[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-github-yalm-3)} ## Kasaysayan ng mga Release | Petsa | Release | Mga Pangunahing Katangian | |--------------|---------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | Hun 2022 | YaLM‑100B | 100 B parameter, 1.7 TB ng datos; Apache 2.0.^{[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-github-yalm-3)} | | 17 Mayo 2023 | YandexGPT 1.0 | Integrasyon sa "Alice".^{[\[1\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-rbc-may17-1)} | | 7 Sep 2023 | YandexGPT 2 | +67% kalidad ayon sa mga panloob na pagsubok.^{[\[4\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-rbc-sep8-4)} | | 28 Mar 2024 | YandexGPT 3 Pro / Lite | Bagong corporate na linya ng API.^{[\[5\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-rbc-mar28-5)} | | 24 Okt 2024 | YandexGPT 4 Pro / Lite | Konteksto ng 32,000 token; nakatagong pangangatwiran (chain‑of‑thought).^{[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-habr-gpt4-6)} | | 25 Peb 2025 | YandexGPT 5 Pro | Pagkakapantay-pantay sa GPT‑4o sa 64% ng mga gawain.^{[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-adindex-5pro-7)} | | 31 Mar 2025 | YandexGPT 5 Lite Instruct | 8‑bilyong modelo sa bukas na access; format ng Llama.^{[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-hf-5lite-8)} | Mga Pangunahing Bersyon ## Arkitektura at Pagsasanay - Pangunahing arkitektura: transformer, na-optimize para sa wikang Ruso. - YandexGPT 5 Lite: Llama‑compatible; pre-training ≈ 15 trilyong token, kasunod na fine‑tuning ≈ 320 bilyon.^{[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-hf-5lite-8)} ### Konteksto at mga Limitasyon - Arkitekturang limitasyon ng konteksto — 32,000 token (mga bersyon 4/5).^{[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-habr-gpt4-6)} - Nililimitahan ng pampublikong API ang isang kahilingan (prompt + completion) sa 7,400 token.^{[\[9\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-langchain-7400-9)} - Pinakamataas na laki ng **sagot** — 2,000 token ayon sa seksyong "Quotas and limits".^{[\[10\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-cloud-limits-10)} ## Kasalukuyang mga Modelo (Hunyo 2025) | Modelo | Parameter | Konteksto | Lisensya | Mga Tala | |------------------|---------------|-----------|-------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------| | YandexGPT 5 Pro | hindi tinukoy | 32,000 | proprietary | Access sa pamamagitan ng API at "Alice Pro".^{[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-adindex-5pro-7)} | | YandexGPT 5 Lite | 8 bilyon | 32,000 | Yandex GPT‑Lite License | Bukas; Llama‑compatible.^{[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-hf-5lite-8)} | | YaLM‑100B | 100 bilyon | 2,048 | Apache 2.0 | Orihinal na proyekto.^{[\[3\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-github-yalm-3)} | ## Mga Benchmark - Mga panloob na pagsubok: ang 5 Pro ay nakamit ang pagkakapantay-pantay sa GPT‑4o sa 64% ng mga gawain; ang paglamang sa 4 Pro — 67%.^{[\[7\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-adindex-5pro-7)} - ru‑LLM Arena: pinapanatili ng YandexGPT ang pamumuno sa ELO‑rating sa mga modelong nakatuon sa wikang Ruso.^{[\[11\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-llmarena-11)} ## Fine‑tuning Para sa 5 Lite, opisyal na sinusuportahan ang paraan ng LoRA; ang halimbawa ng pagpapatakbo ay nailathala sa model card.^{[\[8\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-hf-5lite-8)} ## Mga Mode ng API - Synchronous — mabilis na mga sagot (Lite). - Asynchronous — mga gawaing matrabaho sa resources (Pro).^{[\[2\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-cloud-api-2)} ## Multimodality Ang pamilya ng YandexGPT ay nananatiling text-based; ang mga multimodal na serbisyo ("Neuro", "YandexArt", "Yandex Vision") ay hiwalay na pinauunlad.^{[\[6\]](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_note-habr-gpt4-6)} ## Mga Link - YandexGPT sa Yandex Cloud — pahina ng serbisyo - Mga timbang ng YandexGPT‑5 Lite sa Hugging Face - Repositoryo ng YaLM‑100B sa GitHub ## Literaturang Pinagkunan - 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. ## Mga Tala 1. ↑ ^{[1.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-rbc-may17_1-0)} ^{[1.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-rbc-may17_1-1)} «“Яндекс” добавил в “Алису” аналог ChatGPT». *РБК*. [1] 2. ↑ ^{[2.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-cloud-api_2-0)} ^{[2.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-cloud-api_2-1)} «Getting started with YandexGPT (Quickstart)». *Yandex Cloud Docs*. [2] 3. ↑ ^{[3.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-github-yalm_3-0)} ^{[3.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-github-yalm_3-1)} ^{[3.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-github-yalm_3-2)} «yandex/YaLM‑100B: Pretrained language model with 100B». *GitHub*. [3] 4. [↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-rbc-sep8_4-0) «Как “Яндекс” решил зарабатывать на своём аналоге ChatGPT». *РБК*. [4] 5. [↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-rbc-mar28_5-0) «“Яндекс” представил третье поколение нейросетей YandexGPT». *РБК*. [5] 6. ↑ ^{[6.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-habr-gpt4_6-0)} ^{[6.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-habr-gpt4_6-1)} ^{[6.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-habr-gpt4_6-2)} «Более мощное семейство моделей YandexGPT 4». *Habr*. [6] 7. ↑ ^{[7.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-adindex-5pro_7-0)} ^{[7.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-adindex-5pro_7-1)} ^{[7.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-adindex-5pro_7-2)} «“Яндекс” внедрил YandexGPT 5 Pro в чат с “Алисой Про”». *AdIndex*. [7] 8. ↑ ^{[8.0](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-hf-5lite_8-0)} ^{[8.1](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-hf-5lite_8-1)} ^{[8.2](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-hf-5lite_8-2)} ^{[8.3](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-hf-5lite_8-3)} «yandex/YandexGPT‑5‑Lite‑8B‑pretrain». *Hugging Face*. [8] 9. [↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-langchain-7400_9-0) «ChatYandexGPT API Reference (max_tokens = 7400)». *LangChain Docs*. [9] 10. [↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-cloud-limits_10-0) «Yandex Cloud service quotas and limits → Foundation Models». *Yandex Cloud Docs*. [10] 11. [↑](https://systems-analysis.info/int/YandexGPT_(language_model)_(TL)#cite_ref-llmarena_11-0) «llmarena/llmarena — российская краудсорсинговая платформа оценки LLM». *GitHub*. [11]