gptme

Open-source terminal agent that writes code, runs commands, browses the web, and handles images.

4.5 (6)

Anmeldt av Daniel Nikulshyn·Oppdatert mai 2026

Open Source CLI Developer Tools Coding Agent Local-First Self-Hosted Automation

Oversikt

gptme is a command-line AI assistant that lives in your terminal and acts as a hands-on coding and automation agent. It can read and edit files, execute shell commands, run Python snippets, fetch and browse web pages, and process images, making it useful for everything from quick scripts to longer development sessions. Unlike hosted chat tools, gptme runs locally and works with a range of model backends, including OpenAI, Anthropic, and local models via Ollama or other providers. Conversations are stored as plain files you can resume, edit, or share, and tool use happens transparently in the same terminal where you already work. It is aimed at developers who want a scriptable, transparent alternative to closed coding assistants, with the flexibility to extend tools, pipe input from other commands, and integrate it into existing workflows.

Nøkkelfunksjoner

Shell and Python command execution
File reading and editing in the working directory
Web browsing and page fetching
Vision support for image inputs
Pluggable model backends (OpenAI, Anthropic, local)
Persistent, file-based conversation history

Priser

Modell: Free
Kategori: AI Agents Frameworks
Vurdering: 4.5 / 5 (6)

Brukstilfeller

Automate coding tasks from the terminal

Use gptme to read, edit, and refactor files in your working directory while running shell or Python commands, keeping the entire development loop inside your terminal.

Run a private agent with local models

Pair gptme with Ollama or other local backends to get a hands-on AI assistant that executes commands and edits code without sending data to hosted chat services.

Script repeatable AI workflows

Integrate gptme into existing shell scripts and pipelines to automate web fetching, file processing, or image analysis, with conversations saved as plain files for reuse.

Resume and share development sessions

Because conversation history is stored as editable files, developers can pause long coding sessions, tweak context, or share transcripts with teammates for review.

Fordeler og ulemper

Fordeler

Runs entirely in the terminal with minimal setup
Supports multiple model providers, including local ones
Transparent tool use you can inspect and control
Open source and scriptable into existing workflows

Ulemper

Command-line interface has a learning curve
Capabilities depend heavily on the chosen model
Less polished UI than commercial coding assistants

Anmeldelser

4.5

Gjennomsnitt fra 6 vurderinger.

Logg inn for å legge igjen en anmeldelse.

Carlos Mendoza

Compared a few options

Evaluated this against two competitors. Where it wins: file reading and editing in the working directory and supports multiple model providers, including local ones. Where it lags: command-line interface has a learning curve. On balance the feature set — especially shell and Python command execution — justifies the 4 stars for our use case.

Leila Hassan

Does the job

Pretty happy overall. File reading and editing in the working directory just works and runs entirely in the terminal with minimal setup. Command-line interface has a learning curve can be annoying, but no dealbreakers — I'd recommend it to a friend without hesitating.

Wei Chen

Use it every day

Honestly didn't expect to like it this much. Vision support for image inputs is exactly what I needed, and transparent tool use you can inspect and control. but I reach for it almost every day now and it just clicks.

Omar Haddad

Skeptical, then convinced

I went in skeptical — most tools in this space overpromise. It actually delivers on pluggable model backends (OpenAI, Anthropic, local), and transparent tool use you can inspect and control caught me off guard. still, I'd recommend giving it a real trial.

Tariq Aziz

Years in this space

I've evaluated a lot of these over the years. What stands out here is persistent, file-based conversation history — handled better than most — and runs entirely in the terminal with minimal setup. Worth the time if this is your use case.

Camille Laurent

Years in this space

I've evaluated a lot of these over the years. What stands out here is persistent, file-based conversation history — handled better than most — and transparent tool use you can inspect and control. Command-line interface has a learning curve is my one real gripe. Worth the time if this is your use case.