> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/SuperCmdLabs/SuperCmd/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Providers

> Configure AI integrations for chat, voice input, and text-to-speech

## Overview

SuperCmd supports multiple AI providers for different features:

* **LLM Chat**: OpenAI, Anthropic, Google Gemini, Ollama, OpenAI-compatible
* **Speech-to-Text**: Native macOS, OpenAI Whisper, ElevenLabs
* **Text-to-Speech**: Edge TTS (free), ElevenLabs, Native macOS
* **Memory**: Supermemory integration

All AI settings are configured in **Settings → AI** or by editing `~/Library/Application Support/SuperCmd/settings.json`.

<Note>
  AI features are optional. SuperCmd works perfectly fine without any AI provider configured.
</Note>

## LLM Providers

### OpenAI

**Models supported:**

* `gpt-4o` (recommended)
* `gpt-4o-mini` (faster, cheaper)
* `gpt-4-turbo`
* `gpt-3.5-turbo`

**Setup:**

<Steps>
  <Step title="Get API Key">
    1. Go to [platform.openai.com](https://platform.openai.com)
    2. Create an account or sign in
    3. Navigate to **API Keys**
    4. Click **Create new secret key**
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "enabled": true,
        "provider": "openai",
        "openaiApiKey": "sk-...",
        "defaultModel": "gpt-4o-mini"
      }
    }
    ```
  </Step>

  <Step title="Test">
    Open SuperCmd and type `?hello` to test AI chat
  </Step>
</Steps>

**Pricing (as of 2024):**

* GPT-4o: $5 per 1M input tokens, $15 per 1M output tokens
* GPT-4o-mini: $0.15 per 1M input tokens, $0.60 per 1M output tokens

***

### Anthropic (Claude)

**Models supported:**

* `claude-sonnet-4-20250514` (recommended)
* `claude-3-5-sonnet-20241022`
* `claude-3-5-haiku-20241022`
* `claude-3-opus-20240229`

**Setup:**

<Steps>
  <Step title="Get API Key">
    1. Go to [console.anthropic.com](https://console.anthropic.com)
    2. Create an account
    3. Navigate to **API Keys**
    4. Click **Create Key**
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "enabled": true,
        "provider": "anthropic",
        "anthropicApiKey": "sk-ant-...",
        "defaultModel": "claude-sonnet-4-20250514"
      }
    }
    ```
  </Step>

  <Step title="Test">
    Type `?what is SuperCmd` in the launcher
  </Step>
</Steps>

**Pricing:**

* Claude Sonnet: $3 per 1M input tokens, $15 per 1M output tokens
* Claude Haiku: $0.25 per 1M input tokens, $1.25 per 1M output tokens

***

### Google Gemini

**Models supported:**

* `gemini-2.0-flash-exp` (recommended)
* `gemini-1.5-flash`
* `gemini-1.5-pro`

**Setup:**

<Steps>
  <Step title="Get API Key">
    1. Go to [aistudio.google.com](https://aistudio.google.com)
    2. Click **Get API Key**
    3. Create a new key or use existing
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "enabled": true,
        "provider": "gemini",
        "geminiApiKey": "AI...",
        "defaultModel": "gemini-2.0-flash-exp"
      }
    }
    ```
  </Step>
</Steps>

**Pricing:**

* Gemini Flash: Free tier available, then \$0.075 per 1M tokens
* Gemini Pro: \$1.25 per 1M input tokens

***

### Ollama (Local)

**Models available:**

* `llama3.2` (3B, 1B)
* `qwen2.5` (7B, 14B, 32B)
* `mistral` (7B)
* Any model from [ollama.com/library](https://ollama.com/library)

**Setup:**

<Steps>
  <Step title="Install Ollama">
    ```bash theme={null}
    brew install ollama
    ```
  </Step>

  <Step title="Pull a model">
    ```bash theme={null}
    ollama pull llama3.2
    ```
  </Step>

  <Step title="Start Ollama server">
    ```bash theme={null}
    ollama serve
    ```
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "enabled": true,
        "provider": "ollama",
        "ollamaBaseUrl": "http://localhost:11434",
        "defaultModel": "llama3.2"
      }
    }
    ```
  </Step>
</Steps>

<Info>
  Ollama runs models locally — no API costs, but requires significant RAM and CPU.
</Info>

**System requirements:**

* 8GB RAM minimum (16GB recommended)
* Apple Silicon Mac recommended for best performance
* 4-10GB disk space per model

***

### OpenAI-Compatible APIs

Supports any API that follows the OpenAI chat completions format:

* OpenRouter
* Together AI
* Groq
* Local LLMs (LM Studio, LocalAI)

**Setup:**

```json theme={null}
{
  "ai": {
    "enabled": true,
    "provider": "openai-compatible",
    "openaiCompatibleBaseUrl": "https://api.openrouter.ai/api/v1",
    "openaiCompatibleApiKey": "sk-or-...",
    "openaiCompatibleModel": "anthropic/claude-3.5-sonnet"
  }
}
```

<Tip>
  Use OpenRouter to access multiple models (Claude, GPT-4, Llama) through a single API.
</Tip>

***

## Speech-to-Text (STT)

### Native macOS (Default)

**Free, on-device, no API key required.**

```json theme={null}
{
  "ai": {
    "speechToTextModel": "native",
    "speechLanguage": "en-US"
  }
}
```

**Supported languages:**

* English: `en-US`, `en-GB`, `en-AU`
* Spanish: `es-ES`, `es-MX`
* French: `fr-FR`
* German: `de-DE`
* Italian: `it-IT`
* Japanese: `ja-JP`
* Korean: `ko-KR`
* Chinese: `zh-CN`, `zh-TW`
* And 50+ more via `SFSpeechRecognizer`

**Source:** `src/native/speech-recognizer.swift`

***

### OpenAI Whisper

**Models:**

* `whisper-1` (hosted by OpenAI)

**Setup:**

```json theme={null}
{
  "ai": {
    "speechToTextModel": "whisper-1",
    "openaiApiKey": "sk-..."
  }
}
```

**Pricing:** \$0.006 per minute of audio

***

### ElevenLabs

**Models:**

* `eleven_multilingual_v2`

**Setup:**

```json theme={null}
{
  "ai": {
    "speechToTextModel": "eleven_multilingual_v2",
    "elevenlabsApiKey": "..."
  }
}
```

**Pricing:** Part of ElevenLabs subscription

***

## Text-to-Speech (TTS)

### Edge TTS (Default)

**Free, cloud-based, no API key required.**

```json theme={null}
{
  "ai": {
    "textToSpeechModel": "edge-tts",
    "edgeTtsVoice": "en-US-EricNeural"
  }
}
```

**Popular voices:**

* English: `en-US-EricNeural`, `en-US-JennyNeural`, `en-GB-SoniaNeural`
* Spanish: `es-ES-AlvaroNeural`, `es-MX-DaliaNeural`
* French: `fr-FR-DeniseNeural`
* German: `de-DE-ConradNeural`
* Japanese: `ja-JP-NanamiNeural`

Full list: [npm:edge-tts](https://www.npmjs.com/package/edge-tts)

**Source:** `src/main/ai-provider.ts` (using `node-edge-tts`)

***

### ElevenLabs

**High-quality, natural-sounding voices.**

**Setup:**

<Steps>
  <Step title="Get API Key">
    1. Go to [elevenlabs.io](https://elevenlabs.io)
    2. Sign up for an account
    3. Navigate to **Profile → API Keys**
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "textToSpeechModel": "elevenlabs",
        "elevenlabsApiKey": "..."
      }
    }
    ```
  </Step>

  <Step title="Choose a voice">
    Default voices:

    * `Rachel` (21m00Tcm4TlvDq8ikWAM)
    * `Domi` (AZnzlk1XvdvUeBnXmlld)
    * `Bella` (EXAVITQu4vr4xnSDxMaL)
  </Step>
</Steps>

**Pricing:**

* Free tier: 10,000 characters/month
* Starter: \$5/month for 30,000 characters
* Creator: \$22/month for 100,000 characters

**Testing ElevenLabs:**

```bash theme={null}
curl -X POST "https://api.elevenlabs.io/v1/text-to-speech/21m00Tcm4TlvDq8ikWAM?output_format=mp3_44100_128" \
  -H "xi-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text":"Hello from SuperCmd","model_id":"eleven_multilingual_v2"}' \
  --output test.mp3
```

***

### Native macOS TTS

**On-device, free.**

```json theme={null}
{
  "ai": {
    "textToSpeechModel": "native"
  }
}
```

Uses `NSSpeechSynthesizer` — decent quality but less natural than Edge TTS or ElevenLabs.

***

## Supermemory Integration

**Add long-term memory to AI chat.**

[Supermemory](https://supermemory.ai) stores context from previous conversations and retrieves relevant information automatically.

**Setup:**

<Steps>
  <Step title="Create Supermemory account">
    Go to [supermemory.ai](https://supermemory.ai) and sign up
  </Step>

  <Step title="Get API credentials">
    1. Navigate to **Settings → API**
    2. Copy your API key and client ID
  </Step>

  <Step title="Configure SuperCmd">
    ```json theme={null}
    {
      "ai": {
        "supermemoryApiKey": "...",
        "supermemoryClient": "...",
        "supermemoryBaseUrl": "https://api.supermemory.ai",
        "supermemoryLocalMode": false
      }
    }
    ```
  </Step>
</Steps>

**Environment variable fallbacks:**

```bash theme={null}
export SUPERMEMORY_API_KEY="..."
export SUPERMEMORY_CLIENT="..."
```

**Source:** `src/main/memory.ts`

***

## Speech Correction

**Automatically fix speech-to-text errors using an LLM.**

```json theme={null}
{
  "ai": {
    "speechCorrectionEnabled": true,
    "speechCorrectionModel": "gpt-4o-mini"
  }
}
```

When enabled, transcribed text is sent to the correction model to:

* Fix capitalization
* Add punctuation
* Correct common speech-to-text errors

<Warning>
  Speech correction adds \~500ms latency and uses additional API credits.
</Warning>

***

## Feature Toggles

You can enable/disable individual AI features:

```json theme={null}
{
  "ai": {
    "enabled": true,
    "llmEnabled": true,
    "whisperEnabled": true,
    "readEnabled": true
  }
}
```

| Setting          | Feature                     | Default |
| ---------------- | --------------------------- | ------- |
| `enabled`        | Master AI toggle            | `true`  |
| `llmEnabled`     | AI chat (`?` prefix)        | `true`  |
| `whisperEnabled` | Voice input (hold-to-speak) | `true`  |
| `readEnabled`    | Text-to-speech (Read)       | `true`  |

***

## Troubleshooting

<Accordion title="AI chat doesn't work">
  1. Check that `ai.enabled` is `true`
  2. Verify your API key is correct
  3. Test API key with curl:
     ```bash theme={null}
     curl https://api.openai.com/v1/models \
       -H "Authorization: Bearer YOUR_KEY"
     ```
  4. Check console logs (Cmd+Option+I) for errors
</Accordion>

<Accordion title="Voice input doesn't transcribe">
  1. Ensure microphone permission is granted
  2. Test microphone with Voice Memos app
  3. Check `ai.speechToTextModel` setting
  4. For Whisper: verify `openaiApiKey`
  5. Try switching to `native` model
</Accordion>

<Accordion title="API rate limit errors">
  OpenAI free tier limits:

  * 3 requests per minute
  * 200 requests per day

  Solutions:

  * Upgrade to paid plan
  * Switch to `gpt-4o-mini` (cheaper)
  * Use Ollama (local, no limits)
</Accordion>

<Accordion title="Ollama connection refused">
  1. Ensure Ollama is running: `ollama serve`
  2. Check base URL: `http://localhost:11434`
  3. Test with: `curl http://localhost:11434/api/tags`
  4. Verify model is pulled: `ollama list`
</Accordion>

***

## Next Steps

<CardGroup cols={2}>
  <Card title="Voice Input" icon="microphone" href="/features/voice-input">
    Use hold-to-speak dictation
  </Card>

  <Card title="Text-to-Speech" icon="volume-high" href="/features/text-to-speech">
    Read selected text aloud
  </Card>

  <Card title="AI Integration" icon="brain" href="/features/ai-integration">
    Learn about AI features
  </Card>

  <Card title="Settings" icon="gear" href="/configuration/settings">
    All configuration options
  </Card>
</CardGroup>