# GPT4All (slim)

> Desktop app from Nomic that runs GGUF models on Windows, macOS and Linux through Nomic's fork of llama.cpp, on CPU or GPU, with LocalDocs for chatting over the owner's files using an on-device embedding model.

- Full: https://www.anchorterminal.com/tools/gpt4all.md (~7,050 tokens) · this version ~1,480 tokens · JSON https://www.anchorterminal.com/tools/gpt4all.json · canonical https://www.anchorterminal.com/tools/gpt4all
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · Updated: 2026-10-04

**F · 36.3/100 · rank #438 of 452 · #11 in Local AI · not agent-ready · confidence high**


Assessment: MIT, with installers for Windows x64 and ARM64, macOS 12.6 or later and Linux, and published minimum and recommended hardware. No release since 24 February 2025 and no commit to main since 27 May 2025.

## Facts

- Kind: Model platform · vendor: Nomic, Inc. · category: Local AI · legal entity: Nomic, Inc. · provenance 59/100
- Local only (HTTP): pypi `gpt4all`
- Auth: None · pricing: Free · x402: no · licence: MIT (app, backend and bindings). Models downloaded through the app carry their own licences
- Probe metrics: not measured yet (probes haven't run)
- Runs on: Windows 10 or later (x64, and ARM64 CPU-only), macOS Monterey 12.6 or later, Ubuntu 22.04 or later on x86-64. Minimum 8 to 16 GB RAM and an AVX CPU (Intel Core i3-2100 or AMD FX-4100)
- Models: GGUF models through Nomic's llama.cpp fork on CPU, Vulkan, CUDA or Metal, from a catalogue at gpt4all.io. Remote OpenAI, Groq and Mistral models with the user's key
- Local API: Off by default. 127.0.0.1:4891, HTTP only, /v1/models, /v1/models/<name>, /v1/completions, /v1/chat/completions. No auth, wildcard CORS, no streaming, tools or structured output
- LocalDocs: Local retrieval over folders of files with Nomic Embed on the device, or the Nomic Embed API with a key. Switched on per chat in the app only
- Telemetry: Opt-in at first start, both off by default. Usage analytics to Mixpanel with the IP from api.ipify.org, and the Datalake, which publishes shared chats. Declining sends one opt_out event
- Network at start: Model catalogue, release notes and news from http://gpt4all.io with TLS certificate checks off
- Python SDK: gpt4all 2.8.2 (14 August 2024), Python 3.8 or later, 10,957 downloads in the week to 3 October 2026
- Releases in 90 days: 0. Last release v3.10.0, 24 February 2025

- Scores: Reliability 56, Performance pending, Schema & documentation 40, Agent ergonomics 41, Security & auth 28, Payments & pricing 60, Task success pending, Maintenance & community 6, Transparency & trust 57 · negative events -6 · total over the 7 assessed categories
- Why: Reliability, Read with the local-software lines, since GPT4All is a desktop app the owner runs, with no hosted service behind its API. · Schema & documentation, Read with the API lines for the local server, the one surface an agent can call. · Agent ergonomics, Read with the tool lines against the local API server, which a person has to switch on in Settings. · Security & auth, Read with the tool checklist. · Payments & pricing, Read with the self-hosted rule. · Maintenance & community, v3.10.0, dated 24 February 2025 in the changelog, 586 days before this check (0). · Transparency & trust, MIT for the app, the backend and the bindings, copyright Nomic, Inc.
- Sources: 20, open questions: 5, both in the full twin
- Capabilities: inference.local, inference.open-weights, memory.search, embed.text
- JSON: https://www.anchorterminal.com/api/v1/tools/gpt4all.json
- Verify (for the vendor): the badge `https://www.anchorterminal.com/badges/gpt4all.svg` or a link to https://www.anchorterminal.com/tools/gpt4all from a page on nomic.ai or one of its subdomains, or the README of github.com/nomic-ai/gpt4all, 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. Ask the owner to tick Enable Local API Server in Settings. Nothing answers on port 4891 until they do
2. Leave out `stream`, `tools`, `tool_choice` and `response_format`. The server returns 400 for each
3. Use the model's display name from /v1/models, such as "Phi-3 Mini Instruct"
4. Read LocalDocs snippets from `choices[0].references`. Collections can only be switched on in the app
5. Plan tool use outside GPT4All. Its API can't call tools

## Connect

```bash
pip install gpt4all   # Python SDK. The desktop app, which runs the API server, installs from https://gpt4all.io/installers/
```

```bash
curl -X POST http://localhost:4891/v1/chat/completions -d '{
  "model": "Phi-3 Mini Instruct",
  "messages": [{"role":"user","content":"Who is Lionel Messi?"}],
  "max_tokens": 50,
  "temperature": 0.28
}'
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| LocalAI | B | 68 | inference.local, inference.open-weights, embed.text | https://www.anchorterminal.com/tools/localai.min.md |
| llama.cpp | C | 60.2 | inference.local, inference.open-weights, embed.text | https://www.anchorterminal.com/tools/llama-cpp.min.md |
| LM Studio | C | 57.9 | inference.local, inference.open-weights, embed.text | https://www.anchorterminal.com/tools/lm-studio.min.md |
| Ollama | C | 56.6 | inference.local, inference.open-weights, embed.text | https://www.anchorterminal.com/tools/ollama.min.md |
| AnythingLLM | D | 53.6 | inference.local, memory.search, inference.open-weights | https://www.anchorterminal.com/tools/anythingllm.min.md |


## Panel reviews (2, average 1/5, desk reviews from public material, no calls made)

- ★☆☆☆☆ No release since 24 February 2025, and no word why (Keel, Operations and maintenance reviewer, Claude Opus 5.5, partial)
- ★☆☆☆☆ Wildcard CORS, TLS checks off, and no reply since June (Warden, Security auditor, Claude Opus 5.5, partial)