Devika — Open-Source AI Software Engineer

What it is

Devika is an open-source autonomous AI agent that acts as a software engineer. Given a high-level instruction, it researches the web, creates a plan, and writes code to accomplish the task. It positions itself as an open-source alternative to Devin.

The project targets developers and teams who want an AI pair-programmer that handles end-to-end coding tasks, from research to implementation, without manual step-by-step guidance.

How it saves time or tokens

Devika automates the research-plan-code cycle. Instead of manually searching documentation, drafting an architecture, and writing boilerplate, you give Devika a single instruction and it handles the full loop. This reduces the back-and-forth prompting that typically consumes tokens in interactive LLM sessions.

How to use

1. Clone and install Devika:

git clone https://github.com/stitionai/devika.git
cd devika
uv venv
source .venv/bin/activate
uv pip install -r requirements.txt

1. Start the backend and UI:

python devika.py  # backend
cd ui && bun run dev  # frontend

1. Open the browser UI, describe your task in plain English, and let Devika plan and execute.

Example

# Clone the repo
git clone https://github.com/stitionai/devika.git
cd devika

# Setup environment
uv venv && source .venv/bin/activate
uv pip install -r requirements.txt

# Launch
python devika.py
# In another terminal:
cd ui && bun run dev

Then open http://localhost:3000, enter a task like 'Build a REST API with FastAPI that manages a todo list', and Devika will research FastAPI docs, plan the file structure, and generate the code.

Related on TokRepo

• AI Tools for Coding — More AI-powered coding tools and agents
• AI Tools for Automation — Automation workflows that complement autonomous agents

Key considerations

When evaluating Devika for your workflow, consider the following factors. First, assess whether your team has the technical prerequisites to adopt this tool effectively. Second, evaluate the maintenance burden against the productivity gains. Third, check community activity and documentation quality to ensure long-term viability. Integration with your existing toolchain matters more than feature count alone. Start with a small pilot project before rolling out across the organization. Monitor resource usage during the initial adoption phase to identify bottlenecks early. Document your configuration decisions so team members can onboard independently.

Common pitfalls

• Devika requires API keys for the underlying LLM providers; without them, the agent cannot function.
• Complex multi-repo tasks may exceed context limits, causing incomplete plans.
• The web research step depends on internet access and may fail behind strict firewalls or proxies.