Jakub Ćwirlej

Jakub Ćwirlej is a small 164.5M-parameter code language model that provides autocomplete, next-line suggestions, and code translation.

EUGEN 160M is a decoder-only transformer model designed to assist with writing code. It completes code as you type, suggests the next line, and fills in missing code in the middle of files. The model also specialises in translating Python code into C++, as well as translation pairs involving Java, Go, and Rust. Trained with MuonH on a single NVIDIA RTX PRO 6000 GPU, it features 36 layers and 164.5 million parameters. Its compact footprint enables execution speeds of up to 364 tokens per second on an ordinary CPU.

Jakub Ćwirlej at a glance

Best for: Software engineer
Runs on: CLI
Pricing: Free
Released: October 2026
Category: AI coding tool
Website: jakubcwirlej.com

Key features

Code autocomplete as you type
Next-line code suggestions
Fill-in-the-middle code completion
Python to C++ code translation
CPU generation up to 364 tokens per second
Compiler and test verification for translations
Terminal-based execution

Use it for

Completing code inline while typing
Translating Python programs into verified C++
Filling in missing code blocks in the middle of files
Running code assistance models directly in the terminal

Alternatives to Jakub Ćwirlej

jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
OpenCode — Works with 75+ model providers, local models, and existing subscriptions
LangChain4j — Integrates LLMs and vector stores directly into Java applications.
Strata — Runs large AI models locally on consumer gaming PCs without server reliance.
AirLLM — Runs massive 70B+ parameter models on small 4GB consumer GPUs.
RTK — Reduces AI context output noise by 56 to 90 percent with zero configuration.
16x Prompt — Manage code context and compare multiple LLM responses side-by-side easily.
CodeSight — Reduces AI token usage significantly and maps projects instantly.
SmallCode — Runs fully local with budget-managed context and a forgiving parser for small models.
Muse Wonder — Explore performance benchmarks for Muse Code and Muse Spark.
EasyCode — Answers questions and provides context-aware suggestions inside the integrated development environment.
Deep Code CLI — Optimized specifically for DeepSeek models with a one million token context.

All Jakub Ćwirlej alternatives