AutoResearch vs autoresearch
AI-driven research workflows compared
AutoResearch supports a broad research workflow, from idea discovery or a supplied hypothesis through experiments and independent review. autoresearch focuses on repeated language-model training experiments, with an AI agent changing training code and comparing results under a fixed time budget.
AutoResearch: Turn Research Ideas into Reviewable Evidence
AutoResearch is a Python agent workflow for developing AI and machine learning research ideas into experiments and reviewable evidence. It suits researchers who want a traceable, resumable process and can configure model services and experiment resources.
autoresearch: Let AI Agents Run LLM Training Experiments
autoresearch gives an AI coding agent a small, single-GPU language-model training setup to modify and test autonomously. It suits researchers with supported NVIDIA hardware who want to explore training changes through repeated, time-bounded experiments.
| AutoResearch | autoresearch | |
|---|---|---|
| Language | Python | Python |
| License | Apache-2.0 | — |
| Stars | 2.9k | 97.2k |
| Forks | 256 | 13.5k |
| Last analyzed | Oct 4, 2026 | Oct 3, 2026 |
Key differences
- AutoResearch covers research planning, implementation, execution, analysis, and review; autoresearch centers on modifying and testing language-model training.
- AutoResearch records plans, queues, logs, results, and decisions for resumable work; autoresearch uses a five-minute experiment loop and validation bits per byte to assess changes.
- AutoResearch can use multiple model roles and requires at least three distinct models for its Idea Forge stage; autoresearch directs an AI coding agent to edit `train.py` using instructions in `program.md`.
- AutoResearch requires an experiment environment and resources suited to the work, with GPU use depending on the experiment; autoresearch's main setup requires one NVIDIA GPU.
- Both projects use Python and are not archived. AutoResearch is Apache-2.0 licensed; autoresearch lists 97.2k stars and 13.5k forks, while AutoResearch lists 2.9k stars and 256 forks.
Choose AutoResearch if you…
- need a traceable workflow spanning research ideas, experiments, analysis, and review.
- want to preserve experiment artifacts and negative results so work can be resumed and inspected.
- can configure model services and the tools and resources required by your experiments.
Choose autoresearch if you…
- want an AI agent to iteratively change language-model training code and evaluate results.
- have a single NVIDIA GPU and want experiments run with a consistent five-minute budget.
- want a compact training setup to inspect or adapt for learning and experimentation.
Compare the numbers side by side, or add more projects
This comparison is generated with AI from the OSRepos analyses of both projects. Always check each project's repository and documentation before choosing.