Models & assistants Model or AI service
Bielik 11B v2 focuses on Polish-language text
SpeakLeash and ACC Cyfronet AGH developed Bielik 11B v2 for Polish-language text tasks. The catalog links the 11B v2.3 instruction-tuned checkpoint. [1][2]
- License
- The Bielik-11B-v2.3-Instruct card lists Apache License 2.0. [2]
What the model does
The Bielik 11B v2 technical report was submitted on May 5, 2025. The maintainers separately describe v2.3 Instruct as a merge of the v2.0, v2.1 and v2.2 instruction checkpoints. [1][2]
Why it matters
Polish benchmarks in the report make it possible to judge performance on tasks centered on the language, including linguistic understanding and reasoning. [1][2]
Who can use it
The family is relevant to developers and researchers assessing Polish text generation and instruction following. [1][2]
Key dates
- Authors submit the Bielik 11B v2 technical report [1]
Ways to use it
The v2.3 card distinguishes the combined instruction checkpoint from the individual checkpoints that contributed to it. [1][2]
Model limitations
The paper reports tests for the v2 model family. Its submission date identifies the research report, and the v2.3 documentation identifies that specific checkpoint. [1][2]
Sources and verification
Catalog evidence checked How we verify records
- Primary source SpeakLeash and Cyfronet authors’ Bielik 11B v2 technical report
- Primary source speakleash/Bielik-11B-v2.3-Instruct
- Other source Bielik 11B v2 Technical Report
Related profiles
Models & assistants
GPT-6.2 Astra and Sol
OpenAI
Astra targets demanding projects. Sol targets everyday tasks. Both support text and image input in the API.
Models & assistants
MAI Code 2 and Code 2 Flash
Microsoft AI
Code 2 targets complex software work. Flash targets frequent edits, explanations and routine fixes.
Models & assistants
MAI Realtime 1
Microsoft AI
Users can try a MAI Labs demo. Developers can register for a limited Foundry preview.