★ Reading this for free? Get 25 structured AI courses + per-chapter AI tutor — the first chapter of every course free, no card.Start free in 30 secondsOr own every course: $149 once
Code Generation Model

WizardCoder Python 34B

WizardCoder Python 34B is WizardLM's Evol-Instruct fine-tune of Meta's Code Llama Python 34B, released in August 2023 for Python code generation. It sits at the heavy end of the LLMs you can run locally: the default Ollama build is a 19.05 GB download, so plan on a 24 GB GPU or 32 GB of system RAM (more in the AI hardware guide).

🐍 Python-Specialized📊 73.2% HumanEval🚀 34B Parameters

Published Benchmark Scores

HumanEval pass@1, as WizardLM reports it

WizardCoder Python 34B73.2%
ChatGPT-3.572.5%
Claude 271.2%
GPT-4 (March 2023 figure)67.0%
The comparison WizardLM published with the release on 26 August 2023, as printed on the model card. The same card lists GPT-4 (May 2023) at 88.4%, so read it as a 2023 snapshot, not a ranking against current models.

EvalPlus columns on the same card

HumanEval+ (stricter tests)64.6%
MBPP73.2%
MBPP+ (stricter tests)59.9%
WizardCoder-33B-V1.1, HumanEval / MBPP79.9% / 78.9%
All pass@1. The newer WizardCoder-33B-V1.1 (trained from DeepSeek Coder) is Ollama's wizardcoder:33b, an 18.82 GB download, but it ships under the Microsoft Research licence (non-commercial research use only); the Python 34B is under the Llama 2 licence.
Model Size (Q4_0)
19.05 GB
HumanEval Score
73.2%
Speed Ceiling, RTX 3090
~49 tok/s
936 GB/s ÷ 19.05 GB; a ceiling, not a measurement
Context Window
16,384 tokens

System Requirements

▸
Operating System
Windows 10+, macOS 12+, Ubuntu 20.04+, Docker
▸
RAM
32GB for CPU-only use (19.05 GB of weights plus ~3.2 GB of KV cache at the 16K context Ollama sets)
▸
Storage
20GB+ free (the default Q4_0 download is 19.05 GB)
▸
GPU
24GB card (RTX 3090 or 4090) holds the 19.05 GB file plus ~3.2 GB of KV cache at 16K context; smaller cards split layers onto the CPU
▸
CPU
No core-count minimum; without a GPU, generation speed is limited by memory bandwidth

Python Code Generation Setup

1

Install Ollama

Get Ollama for local AI development

$ curl -fsSL https://ollama.com/install.sh | sh
2

Pull WizardCoder Python 34B

The Q4_0 build, 19.05 GB (wizardcoder:34b-python-q4_K_M is 20.22 GB)

$ ollama pull wizardcoder:34b-python
3

Test Python Generation

Test Python code generation capabilities

$ ollama run wizardcoder:34b-python "Create a FastAPI CRUD application with SQLAlchemy"
4

Call It from Python

The official ollama package for Python talks to the local Ollama server

$ pip install ollama && python -c "import ollama; print(ollama.generate(model='wizardcoder:34b-python', prompt='Write a Python function that validates an email address')['response'])"

WizardCoder Python 34B Architecture

WizardCoder Python 34B technical architecture showcasing specialized Python training, transformer-based design, and deployment options for code generation workflows

👤
You
💻
Your ComputerAI Processing
👤
🌐
🏢
Cloud AI: You → Internet → Company Servers

First Run and Prompt Format

Terminal
$ollama pull wizardcoder:34b-python
pulling manifest pulling c8ad30822293: 100% ▕██████████████████▏ 19 GB pulling 2d836d77287d: 100% ▕██████████████████▏ 61 B pulling 8fadb9ad1206: 100% ▕██████████████████▏ 106 B pulling bf6237d30a42: 100% ▕██████████████████▏ 30 B pulling a168bedb9a09: 100% ▕██████████████████▏ 456 B verifying sha256 digest writing manifest success
$_

WizardCoder was trained on an Alpaca-style instruction prompt. The Ollama tag ships that template and its system line, so a plain ollama run prompt is wrapped for you; if you load the weights in another runtime, wrap the request yourself:

Below is an instruction that describes a task. Write a response that appropriately completes the request.

### Instruction:
{instruction}

### Response:

Python Development Applications

🔬 Data Science & Analytics

  • • Advanced pandas data manipulation
  • • Machine learning pipelines (scikit-learn, TensorFlow)
  • • Statistical analysis with NumPy/SciPy
  • • Data visualization (matplotlib, seaborn, plotly)
  • • Jupyter notebook automation
  • • Database ETL processes

🌐 Web Development

  • • FastAPI and Django REST APIs
  • • Flask microservices architecture
  • • Async web applications
  • • Database ORM implementations
  • • Authentication and security
  • • WebSocket real-time features

🤖 AI & Machine Learning

  • • Neural network architectures
  • • Computer vision with OpenCV
  • • Natural language processing
  • • MLOps and model deployment
  • • Deep learning with PyTorch
  • • Reinforcement learning algorithms

⚡ Automation & Scripting

  • • System administration scripts
  • • Web scraping and automation
  • • File processing and manipulation
  • • API integration and testing
  • • DevOps and CI/CD pipelines
  • • Task scheduling and monitoring

Build Real AI on Your Machine

RAG, agents, NLP, vision, and MLOps - chapters across 25 courses that take you from reading about AI to building AI.

🎯
AI Learning Path

Go from reading about AI to building with AI

25 structured courses. Hands-on projects. Runs on your machine. Start free.

Or own it for life — Lifetime $149 $599, pay once
Barsha Mohapatra

Written by

Barsha Mohapatra

Founder and Editor, Local AI Master · Founder, Little AI Master

Founder of Little AI Master and the editor behind every course and guide on Local AI Master. Years of tutoring students taught her where AI education actually fails.

📅 Published: 2025-10-25🔄 Last Updated: 2026-11-02✓ Manually Reviewed
Reading now
Join the discussion

Continue Learning

Explore more code generation models and Python AI tools to enhance your development workflow:

More on AI Models Directory
See the full AI Models Directory guide.
📚
Free · no account required

Grab the AI Starter Kit — career roadmap, cheat sheet, setup guide

No spam. Unsubscribe with one click.

🎯
AI Learning Path

Found your model? Now build something with it.

25 hands-on courses — RAG, agents, fine-tuning — all running locally. First chapter free, no card.

Or own it for life — Lifetime $149 $599, pay once
Free Tools & Calculators