Skip to content
SPECIFICTION
Get listed

Scalaris Atlas

scalarislab.com

Free plan WindowsmacOSLinux
Visit scalarislab.com
Spec score 7.3 #16 of 24 in AI Computer Use Agents

What it is

Scalaris Atlas is a local deployment that turns a computer into a private host for AI models and serves them to compatible clients. It accepts OpenAI-compatible clients at localhost and exposes routes for chat completions, embeddings, and model lists. Atlas includes Pulsar, while its model manager lets users install their own models. The API covers a strict subset of the OpenAI HTTP surface; unimplemented endpoints return a typed 501 response. Models are served locally, and the maker says nothing leaves the machine. Egress is denied by default, and non-loopback binds are refused unless explicitly enabled. Atlas has no telemetry or phone-home behavior, according to the maker. Scalaris Toolbox installs and supervises it. Configuration is kept in one TOML file and reloads when changed. GPU exhaustion can trigger CPU fallback only with --allow-cpu. It is free forever at free and runs on macOS, Linux, and Windows.

Spec sheet

What it does
Atlas turns a computer into a private AI host that serves local models to compatible clients.Read 8 Oct 2026 · scalarislab.com
Intended users
Atlas is for people using OpenAI-compatible clients who want to run AI models locally.Read 8 Oct 2026 · scalarislab.com
API compatibility
Atlas accepts OpenAI-compatible clients at localhost and exposes chat completions, embeddings, and model-list routes.Read 8 Oct 2026 · scalarislab.com
API limits
Atlas exposes a strict subset of the OpenAI HTTP surface, and unimplemented endpoints return a typed 501 response.Read 8 Oct 2026 · scalarislab.com
Configuration
Atlas configuration is stored in one TOML file, and settings reload when that file changes.Read 8 Oct 2026 · scalarislab.com
Installation
Atlas is installed and supervised through Scalaris Toolbox.Read 8 Oct 2026 · scalarislab.com
Integration
The maker says to point an OpenAI-compatible client at localhost:8787 to integrate Atlas.Read 8 Oct 2026 · scalarislab.com
Local processing
The maker says models are served locally and nothing leaves the machine.Read 8 Oct 2026 · scalarislab.com
Model support
Atlas includes Pulsar and lets users install their own models through its model manager.Read 8 Oct 2026 · scalarislab.com
Network controls
Egress is denied by default, and Atlas refuses non-loopback binds unless the user explicitly changes the setting.Read 8 Oct 2026 · scalarislab.com
Resource handling
When GPU memory is exhausted, Atlas falls back to CPU only if the user enables --allow-cpu.Read 8 Oct 2026 · scalarislab.com
Telemetry
The maker says Atlas has no telemetry and does not phone home.Read 8 Oct 2026 · scalarislab.com

Plans and prices

Free Free Free forever · zero egress by default Read 8 Oct 2026 · scalarislab.com

Every Scalaris Atlas plan and price →

Strong points and limits

Strong points

  • Free forever at free
  • OpenAI-compatible chat, embedding, and model-list routes
  • Models and processing stay on the local machine
  • Runs on macOS, Linux, and Windows
  • Configuration reloads from one TOML file

Limits

  • API implements only a strict OpenAI subset
  • CPU fallback requires enabling --allow-cpu
  • Installation and supervision require Scalaris Toolbox

Who it suits

Atlas suits people using OpenAI-compatible clients who want local model serving on macOS, Linux, or Windows. It is also suited to users who need API access with local processing and restrictive network defaults.

Verdict

Scalaris Atlas combines local model serving with an OpenAI-compatible API and explicit network controls. Its free plan and broad desktop platform support fit users who accept the implemented API surface and Toolbox-based installation.

Where Specifiction places it

Alternatives to Scalaris AtlasAll alternatives →