# Axolotl (slim)

> Open-source command-line tool and Python package for fine-tuning open language models from one YAML config, covering LoRA, QLoRA, full fine-tuning, preference tuning and GRPO on the owner's GPUs.

- Full: https://www.anchorterminal.com/tools/axolotl.md (~5,250 tokens) · this version ~1,330 tokens · JSON https://www.anchorterminal.com/tools/axolotl.json · canonical https://www.anchorterminal.com/tools/axolotl
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-09

**B · 64.8/100 · rank #307 of 842 · #2 in Fine-tuning · not agent-ready · confidence medium**


Assessment: Axolotl runs a whole fine-tuning job from one YAML file and ships a JSON Schema of its config plus bundled agent docs. It is 0.x software with telemetry on by default, no terms or privacy policy, and the owner supplies the GPU.

## Facts

- Kind: Agent framework · vendor: Axolotl AI · category: Fine-tuning · legal entity: not named · provenance 36/100
- Packages: pypi `axolotl`, oci `axolotlai/axolotl`
- Auth: None · pricing: Free · x402: no · licence: Apache-2.0
- Probe metrics: not measured yet (probes haven't run)
- Interface: CLI (`axolotl train`, `preprocess`, `evaluate`, `inference`, `merge-lora`, `quantize`, `export`, `vllm-serve`) driven by one YAML config
- Methods: SFT, continued pretraining, LoRA, QLoRA, full fine-tuning, QAT, DPO, IPO, KTO, ORPO, GRPO, GDPO, reward modelling
- Runs on: Linux with NVIDIA (Ampere or newer) or AMD GPUs, Python 3.12 or later, PyTorch 2.13 or later. Multi-GPU and multi-node with FSDP2 or DeepSpeed
- Weights: The owner's. Adapter, merged model, quantised model or GGUF, or pushed to the Hugging Face Hub
- Agent docs: `axolotl agent-docs [topic]` and `axolotl config-schema [--field name]`, both offline
- Remote compute: Docker image for RunPod, Vast.ai, Modal and others; guides for Hugging Face Jobs, SkyPilot and Nebius Serverless Jobs; Tinker-compatible APIs through a plugin
- Telemetry: On by default, to PostHog. `AXOLOTL_DO_NOT_TRACK=1` turns it off
- Free tier: All of it

- Scores: Reliability 64, Performance pending, Schema & documentation 80, Agent ergonomics 60, Security & auth 52, Payments & pricing 60, Task success pending, Maintenance & community 88, Transparency & trust 56 · total over the 7 assessed categories
- Why: Reliability, Scored on the local-software lines. · Schema & documentation, Read as a framework an agent drives through a CLI. · Agent ergonomics, Read as a CLI. · Security & auth, Read as local software. · Payments & pricing, Scored by the self-hosted rule. · Maintenance & community, Version 0.20.0 on PyPI on 30 September 2026 (30). · Transparency & trust, Apache-2.0 in the repository, though the PyPI metadata carries no licence field (30).
- Sources: 15, open questions: 6, both in the full twin
- Capabilities: finetune.sft, finetune.lora, finetune.preference, finetune.rl, finetune.export
- JSON: https://www.anchorterminal.com/api/v1/tools/axolotl.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/axolotl.svg` or a link to https://www.anchorterminal.com/tools/axolotl from a page on axolotl.ai or one of its subdomains, or the README of github.com/axolotl-ai-cloud/axolotl, then `POST https://www.anchorterminal.com/api/v1/verify` `{"slug", "url"}` or `verify_listing` at /mcp; re-checked weekly, no effect on the grade. Snippets in the full twin.

## Before you call it

1. Set `AXOLOTL_DO_NOT_TRACK=1` before any command, or training waits 10 seconds and sends usage events to PostHog
2. Run `axolotl agent-docs` and `axolotl config-schema --field <name>` before writing a config; both work offline from the installed package
3. Install torch first, then `uv pip install --no-build-isolation axolotl[deepspeed]`, on Python 3.12 or later with PyTorch 2.13 or later
4. Take example configs from the same release tag as the installed version; minor releases remove and rename config keys
5. Resume an interrupted run with `axolotl train config.yml --resume-from-checkpoint <path>`, then `axolotl merge-lora` and `axolotl export` only when shipping

## Connect

```bash
uv pip install torch==2.14.0 torchvision && uv pip install --no-build-isolation axolotl[deepspeed]   # or: docker run --gpus '"all"' --ipc=host --rm -it axolotlai/axolotl:main-latest
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Fireworks AI Fine-tuning | C | 59 | finetune.sft, finetune.preference, finetune.rl, finetune.lora, finetune.export | https://www.anchorterminal.com/tools/fireworks-fine-tuning.min.md |
| Unsloth | D | 51.5 | finetune.sft, finetune.preference, finetune.rl, finetune.lora, finetune.export | https://www.anchorterminal.com/tools/unsloth.min.md |
| Tinker | D | 51 | finetune.sft, finetune.preference, finetune.rl, finetune.lora, finetune.export | https://www.anchorterminal.com/tools/tinker.min.md |
| Vertex AI Gemini tuning | B | 64.2 | finetune.sft, finetune.preference, finetune.rl, finetune.lora | https://www.anchorterminal.com/tools/vertex-ai-tuning.min.md |
| Microsoft Foundry fine-tuning (Azure OpenAI) | C | 61.1 | finetune.sft, finetune.preference, finetune.rl, finetune.lora | https://www.anchorterminal.com/tools/azure-foundry-fine-tuning.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)