Hey,

Just released a free beta version of my VST plugin Obsidian Neural for a limited amount of users.
Download it freely here:
https://obsidian-neural.com/local.php

Everything runs on the CPU, with Stable Audio 3 Medium. VST/AU/Standalone
Don't need any GPU.

You don't like it?
I don't care.
I don't like you too.

You like it?
Then fuck it.

 

Is this what they call "aggressive marketing"?

 

Subscription based for a single plugin? That's bold.

 

Anyone who writes something like that is either slightly mentally disturbed or somewhat insane.

 

Is insanity contagious? Don't want to find out. Subscription for just one plugin? That's just a hard no.

 

Yep you're right.
Even my psychiatrist doesn't want of me anymore.
That's why I'm here.

Thanks for your "no".
I prefer honesty than indifference.

Yep.

 

Last edited by a moderator: Jul 10, 2026 at 9:33 PM

what extra does it do compared to free wan2gp with Stable Audio 3 Medium?

 

What OBSIDIAN Neural does
Type words → Get musical loops in ~30s. No stopping your creative flow.

Performance
8-track sampler with MIDI triggering (C3-B3)
4 pages per track (A/B/C/D) - Switch variations instantly
8 sequences per page - 256 total patterns for complex live sets
16-step sequencer with multi-measure support
Quantized page changes - Seamless transitions locked to measure boundaries
4 pair crossfaders with master bypass - Blend each deck A/B pair independently with model-aware color morphing, or toggle the entire crossfader section off for direct routing
Plug-and-play MIDI mapping - Auto-configured for the companion mobile controller, with bidirectional feedback (LED states, knob positions) on dedicated MIDI channels. To use: open the MIDI panel (piano icon, bottom-right) → Load Default Mapping.
MIDI learn on every parameter - Map any control to any hardware override, with persistent user mappings
Sound design
Per-page ADSR envelope - Shape the dynamics of each variation independently, editable directly on the waveform
Per-track gain control - Adjust each sample's level (-12 / +12 dB) before mixing, with visual waveform feedback
Per-track multi-mode filter - LP/HP/BP with 12 or 24 dB slopes, drive, cutoff and resonance for sculpting each voice
Per-track 8-band graphic EQ + master EQ - Independent frequency shaping on every voice and global bus, from 40 Hz to 15 kHz
Per-track compressor + master compressor - Full dynamic control on each voice and on the master bus, with threshold, ratio, attack, release and makeup gain
Per-track limiter + master limiter - Peak protection on every voice and on the master bus with adjustable release and ceiling
Per-track distortion - 6 character modes (Soft, Hard, Tube, Fold, Diode, Cubic) with pre/post gain and high-pass cutoff
Per-track chorus - Modulation effect with rate, depth, delay, feedback and dry/wet control for added stereo width
Tempo-synced delay send - 8 time divisions (1/16 to 2 bars), Stereo / Ping-Pong / Mono modes, per-track send level
Reverb send - Per-track reverb with size, damping, width and mix controls
Airwindows Console6 master bus - Analog-modeled saturation for cohesive mix glue
AI generation
Prompt bank with editor - Build, organize and reuse your prompts with model-aware keywords (genres, elements, moods, negatives)
Drag-and-drop prompts - Drop a prompt on a track to assign both prompt and AI model in one gesture
Sample bank with drag-and-drop - Every generation is automatically saved and can be reused across tracks and projects
LLM bypass mode - Skip prompt enhancement for faster generation when you already know what you want
Non-blocking generation - No pre-recorded samples, renders in background
Multi-model engine
In server mode, OBSIDIAN Neural offers 9 specialized AI engines - assign a different one to each track for its strengths:

stable-audio-open-1.0 - Versatile foundation, drums and full-mix textures (80–160 BPM)
Stable Audio 3 Medium - Next-gen flexible full tracks, isolated stems, FX (80–160 BPM) (also the local engine)
Foundation-1 - Tag-based melodic and harmonic phrasing (100–150 BPM)
Audialab EDM Elements - High-energy EDM leads, supersaws, plucks (100–150 BPM)
RC Infinite Pianos - Grand and electric piano performances (100–150 BPM)
RC Vocal Textures - Choral, operatic and atmospheric vocals (100–150 BPM)
SAO Instrumental - Melodic trap, lofi jazz rap, indie stems (75–160 BPM)
StableBeaT - Trap beats and 808 grooves (75–160 BPM)
gluten_v1 - Loopable melodic trap and wavy motifs (90–160 BPM)
The Local Edition runs Stable Audio 3 Medium on your CPU. The other 8 engines are available in server mode.

⚠️ AI generation can produce unexpected results. Feedback welcome on Issues or Discussions.

DAW & Standalone
Automatic tempo sync - Incoming samples are BPM-detected with MiniBPM and time-stretched on load via Signalsmith Stretch to lock to your host tempo, with zero pitch drift
Standalone version - Run OBSIDIAN Neural without a DAW, with built-in transport and tempo control
Ableton Link (Standalone) - Bidirectional network sync of tempo and start/stop with any Link-enabled app
OBSIDIAN Neural is NOT a song generator like Suno or Udio. It's a performance tool: you build your track loop by loop, you're the composer, AI is your loop generator.

 

But use wan2gp if you prefer.

 

well done! You really took this thing far. When i first tried obsidian a few months back it was simply not working for me. This version is really smoothly executed. A little clunky at times, but overall well thought out, and even fun to use! I would love to see the "DJ" thing leaned into more. More effects and a dedicated clip launcher, overall tightening up of things, making it smoother for live use. Congrats on the release, and thanks for the hookup!

 

Thanks man! Means a lot to me! Very happy that the work I did on it these last month makes it better now! That was exhausting sometimes, so thanks!
Don't hesitate to give some feedbacks via the contact page on https://obsidian-neural.com I really want the community to have something solid between their hands.
I keep your suggestions in my to do

 

Who are the "they" you are talking about?

 

I thought you didn't care

 

You understood I didn't care about what? When did I use the "I don't care" expression in my original post?

 

Its literally right there at the top, you said

"You don't like it?
I don't care.
I don't like you too."

 

Yep you're right. I don't care if you don't like my VST.
But I never said I didn't care about why you say what you say. Then I am interested who is the "they" in his affirmation.

 

I bet you smell funny.

 

And you win! How much did you bet?

 

Hey,

really like the VST, and I too would like future developpement lean towards live Dj style/producing and playing live performances...One thing though: is it possible to download the llm model on a internet connected computer, then transfer it to a totally offline machine?

Edit: hmmm no offline activation, can't use it in my live rig, so can't use it for what I intended then.

 

Last edited: Jul 11, 2026 at 7:04 PM

Thanks for your message.
Yep when you open the plugin you can activate the plugin with the License code you received after the download and installation of the beta version. Then after it will download the models.
When did you download the VST for the last time?