Settings Tab

Settings The plumbing. A Common strip that applies everywhere, then three sub-tabs: AI Services, Audio, and Push To Talk.

Common

Shown above the sub-tabs, because it applies to all of them.

Language — the language of both your voice commands and the app's own interface. Choosing one re-renders the whole window immediately and Vega announces the change out loud.

Supported: English, Spanish, French, German, Italian, Portuguese, Brazilian Portuguese, Ukrainian, Russian.

Journal Directory — where Elite Dangerous writes its journal files. Optional: leave it blank and the standard location for your platform is used. This is how Elite Intel knows what is happening around your ship, so if it is wrong the app is effectively blind, and it will say so at startup.

AI Services

AI services

Rewritten in V1.1. The old scattered "Use" checkboxes are gone. There are now two switches — one for the language model, one for speech — and the unused side of each is dimmed so it is obvious which one is live.

This is also the one tab in the app that works on a draft. Nothing is written until you press Save, and trying to leave with unsaved edits asks you to Save, Discard, or Keep editing.

Language Model (LLM)

Switch between Local Setup and Cloud Setup.

Local Setup

Field Notes
Address Defaults to LM Studio's usual URL. Point it at another machine's IP if inference runs elsewhere on your LAN
Model The model name. One field — V1.1 uses a single model for both commands and queries

The default and recommended local model is google/gemma-4-e4b. Elite Intel warns you at startup if your local model is something else; other models may work poorly or not at all.

Setup guides: LM Studio on Linux · LM Studio on Windows · AMD RX series

Cloud Setup

One field: your API Key, with a Locked checkbox beside it so a saved key cannot be edited by accident. Uncheck Locked to change it.

Supported providers: Gemini, Grok, OpenAI, Claude, DeepSeek, Mistral.

You no longer pick a model. Elite Intel recognises the provider from the shape of your key and selects the right model itself.

Mistral has a free tier and is the easiest way to start. See Cloud LLM options for how to obtain a key from each provider.

Speech (TTS)

Switch between Local · Kokoro and Cloud · Google.

  • Local · Kokoro has no configuration at all. 53 voices, built in, no key, no download.
  • Cloud · Google needs a Google TTS Key, with the same Locked checkbox.

Switching engines resets every ship's voice to the new engine's default voice. Ship personalities are kept. You are asked to confirm before it happens.

Footer

Restore Defaults resets the language-model configuration to local LM Studio with the default model, and saves immediately. Save commits everything else; it is greyed out until something actually changes, and an Unsaved changes hint appears next to it when it does.

Saving restarts only what it has to — changing the model restarts the brain, changing the speech key restarts the mouth.

Audio

Audio settings

Audio Devices

Mic and Speaker dropdowns, or (System Default). The same pickers are available from the Audio Devices button on the Vega tab.

Device changes take effect on the next service start.

Enable Noise Reduction with a Low / Medium / High strength. Start at Medium. High is for genuinely noisy rooms — it is aggressive, and over-filtering can cost you transcription accuracy.

Audio Levels

Slider What it does
Speech Volume How loud Vega speaks
TTS Voice Speed How fast Vega speaks
Beep Volume The confirmation beep — it fires when speech-to-text has finished and the language model has your input
STT Threads CPU threads for transcription (4–11). A minimum request, not a reservation: the app asks for this many, uses what the processor gives it, and releases them when the work is done

Microphone Monitor

A live meter down the right side. Three things to read on it:

  • FLOOR — your noise level when you are not speaking.
  • GATE — the threshold. Audio above the gate is streamed for transcription; when it drops below, what was captured is transcribed and sent to the language model.
  • CLIP — you are overdriving the microphone. Anything above this line transcribes badly.

You want a clear gap between FLOOR and your speaking level, and nothing touching CLIP. If that is not what you see, run CALIBRATE AUDIO on the Vega tab — it sets the gate for you, and warns you if the speech-to-noise gap is too small to work with.

Push To Talk

Push to talk

Push-to-talk works with a game controller or HOTAS button, not a keyboard. You give up one button and you gain a microphone that is closed except when you want it open.

Control Notes
Enable Push to Talk The master switch. Everything else is disabled until this is on
Controller Any connected controller Elite Intel can see. It re-selects your saved controller automatically when it reconnects
Button Which button on it

Two modes:

  • Toggle to sleep / wake — the button flips Vega between sleeping and listening. While asleep, Vega ignores everything except Wake up!, and the listen / listen up bypass word still gets a single command through: "Listen up — lower the landing gear."
  • Push to Talk — Vega ignores everything by default. Hold the button, hear a beep, speak, release. A second beep confirms your input is being processed.

While push-to-talk is active, the SLEEP / WAKE UP button on the Vega tab is disabled — the controller button is the gate.

The button works whether or not you ever open this tab.

Where the settings live

All settings and data are stored in a SQLite database:

  • Linux: ~/.local/share/elite-intel/elite-intel/db/
  • Windows: %APPDATA%\elite-intel\db\

Community 👉Matrix👈