acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜

Building MLX Server Manager: an open-source macOS GUI for running mlx_lm.server on Apple Silicon.

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

I’ve made MLX Server Manager public.

It’s an open-source macOS GUI for managing pure mlx_lm.server on Apple Silicon in Direct Mode, making it easier to run local MLX models as OpenAI-compatible endpoints.

https://t.co/akIgxYdQei

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.9.0 is now available.

This app release adds Import Profiles Preview.

You can now select an exported profile JSON file and inspect it before any future import action.

Preview shows:

schemaVersion / app / exportedAt
profile counts
valid / invalid entries
warnings
conflicts with existing profiles
Advanced Launch Options status

This is preview-only.

It does not:

save profiles
replace profiles
rename profiles
start / stop / restart servers
call /v1/models
make external HTTP requests
import API keys, tokens, secrets, executable paths, model weights, or caches

Export Profiles is already available. Actual Import Profiles remains future work.

https://t.co/C8QctJIgUI

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.8.0 is now available.

This docs-only release polishes the future Import Profiles design.

It defines:

Import Preview sheet
validation severity: error / warning / info
unsupported schemaVersion handling
duplicate / conflict handling
Advanced Launch Options validation
side-effect-free import preview boundary

Import Profiles is still future work.

The design keeps import limited to profile metadata:

no model weights
no caches
no API keys / tokens / secrets
no executable paths
no automatic server start
no external process ownership changes

https://t.co/q4roGb6TLk

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.7.0 is now available.

This app release adds Export Profiles.

You can now export model profile metadata to JSON for backup or migration preparation.

Export includes:

profile name
model ID
host / port
non-empty Advanced Launch Options

Export does not include:

API keys / tokens / secrets
executable paths
model weights / caches
logs or runtime state

Export is side-effect-free:

no server start / stop / restart
no external HTTP requests
no /v1/models call
no process ownership changes

Import Profiles remains future work.

https://t.co/Be40CQifDF

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.6.0 is now available.

This docs-only release adds the Model Profile Import / Export design.

The design focuses on safe profile metadata backup, migration, and sharing.

It explicitly excludes:

model weights
caches
API keys / tokens / secrets
executable paths by default
automatic server start

Direct Mode remains unchanged.

https://t.co/thcOcNW9tm

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.5.0 is now available.

This docs/public polish release adds a README screenshot for the First-run Onboarding Guidance panel introduced in v2.4.0.

The screenshot shows:

setup hints for first-time users
Not Running / Not Connected guidance
Direct Mode reminder
no inference proxy behavior

No app code changed. Current binary remains v2.4.0.

https://t.co/4436lUOgJa

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.4.0 is now available.

This release adds a lightweight First-run Onboarding Guidance panel.

It helps new users understand what to configure next:

mlx_lm.server executable path
model profile
managed vs external/adopted server flow
Connection Settings
Direct Mode

No wizard. No auto install. No model downloader. No inference proxy.

https://t.co/NKjOqwWuuI

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.3.0 is now available.

This docs-only release adds First-run Onboarding Guidance.

It documents the initial setup flow for:

mlx_lm.server executable path
model profiles
/v1/models readiness checks
Connection Settings
Hermes Agent setup
managed vs external/adopted server workflows

Direct Mode remains unchanged.

https://t.co/kYomjaZFuA

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.2.0 is now available.

This release refreshes the README screenshots for the current UI.

The new screenshots cover:

Main dashboard
Connection Settings / Current Target summary
Adopted External Server state

They show the current Direct Mode workflow and the external server ownership boundary more clearly.

https://t.co/x30BqW9RSV

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.1.0 is now available.

This docs-only release adds a Screenshot Refresh Design.

It defines the next README screenshot set for:

Main dashboard
Connection Settings / Current Target summary
External Server Detected
Adopted External Server
Advanced Launch Options

It also includes a privacy checklist to avoid exposing tokens, API keys, private paths, or local user info.

https://t.co/bF5DZcyWZz

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v2.0.0 is now available.

This is a docs-only public polish release.

The README and docs now reflect the current feature set:

managed mlx_lm.server
External Server Detection
Adopt External Server
Connection Settings Current Target summary
Hermes Agent / OpenAI-compatible client setup

Direct Mode remains unchanged:
OpenAI-compatible client -> server -> MLX model
https://t.co/SDkomxXfks

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.9.0 is now available.

This release improves the Connection Settings UI.

New Current Target summary shows:

Managed Server
External Server Detected
Adopted External Server
Not Running / Not Connected

It also adds clearer copy actions for Base URL, Model ID, Hermes Agent config, and /v1/models readiness checks.

Direct Mode remains unchanged.

https://t.co/94hB1rveIn

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.8.0 is now available.

This docs-only release adds the Connection Settings Polish design.

The goal is to make the active connection target clearer across:

Managed Server
External Server Detected
Adopted External Server
Not Running

Direct Mode remains unchanged.

https://t.co/hMIE9s98RA

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.7.0 is now available.

This release adds the initial Adopt External Server implementation.

When an external OpenAI-compatible server is detected, you can now explicitly adopt it as a connection context.

Adopt does not mean process ownership:

no Stop / Restart
no PID ownership
no external process kill
no memory/log collection

Direct Mode remains unchanged.

https://t.co/PBs9BW4rhn

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

まだまだ機能は貧弱ですが、「生のmlx-lmに手を入れない、あくまでフロントエンド」をモットーにコツコツと開発していきます。

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MacでローカルLLMを利用する方向けに、mlx-lmをGUIで簡単に管理できるOSSを開発中です。mlx-lmは他と比較して多くの場面で最速トークン生成スピードを誇る一方、UIがCLIのみなので敬遠していた人も多いのでは…。私もその1人で、それじゃあ自分で作ってしまえ、というのがこのPJの始まりです。

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

I’ve made MLX Server Manager public.

It’s an open-source macOS GUI for managing pure mlx_lm.server on Apple Silicon in Direct Mode, making it easier to run local MLX models as OpenAI-compatible endpoints.

https://t.co/akIgxYdQei

#MLX #LocalLLM #AppleSilicon #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.6.0 is out.

This release documents the design for a future Adopt External Server workflow.

The safety boundary is explicit:

no process ownership
no Stop/Restart for external servers
no PID management
no proxying
Direct Mode remains unchanged

https://t.co/xSE8hMe3HK

#MLX #LocalLLM #SwiftUI #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

@NousResearch
@Teknium
Hi! I’m a fan of Hermes Agent.

I’ve been building MLX Server Manager, a macOS GUI for managing mlx_lm.server, to make it easier to use Hermes Agent with local MLX models on Apple Silicon.

I’m a beginner, but I’d love to contribute to the Hermes Agent, Mac, and local LLM communities.

https://t.co/7MsxZJMbPw

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.5.0 is out.

It now detects an already-running OpenAI-compatible server on the selected host/port using /v1/models.

External servers are detected, not taken over:

no Stop/Restart
no PID ownership
no process killing
Direct Mode unchanged

https://t.co/b42umfunNa

#MLX #LocalLLM #SwiftUI #OpenSource

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

@JohnnyNel_ Watching Fable5 convinced me that our future lies in local LLMs.

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

1/2

MLX Server Manager v1.3.0 is now available.

This docs-only release adds a Hermes Agent / OpenAI-compatible client connection guide.

https://t.co/pieDJAHguT

#MLX #LocalLLM #AppleSilicon

[acari | 💵, 🎮, 🥽, 💻, 🥊, 🍜 @Crypto_is_all]

MLX Server Manager v1.4.0 is out.

This release documents the design for future external server detection:

detect /v1/models on the selected host/port
distinguish managed vs external servers
avoid killing external processes
keep Direct Mode unchanged

https://t.co/xbzq280WPV

#MLX #LocalLLM #SwiftUI #OpenSource