Skip to content

Add multi-provider support with model selection and custom model input #43

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
xROBERTINOx wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Add multi-provider support with model selection and custom model input #43

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
276 changes: 276 additions & 0 deletions Sources/APIProvider.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,276 @@
import Foundation

// MARK: - API Provider

enum APIProvider: String, CaseIterable, Identifiable, Codable {
case groq
case openai
case togetherAI = "together_ai"
case fireworks
case nvidia
case custom

var id: String { rawValue }

var displayName: String {
switch self {
case .groq: return "Groq"
case .openai: return "OpenAI"
case .togetherAI: return "Together AI"
case .fireworks: return "Fireworks"
case .nvidia: return "NVIDIA NIM"
case .custom: return "Custom"
}
}

var defaultBaseURL: String {
switch self {
case .groq: return "https://api.groq.com/openai/v1"
case .openai: return "https://api.openai.com/v1"
case .togetherAI: return "https://api.together.xyz/v1"
case .fireworks: return "https://audio-prod.api.fireworks.ai/v1"
case .nvidia: return "https://integrate.api.nvidia.com/v1"
case .custom: return ""
}
}

var availableTranscriptionModels: [TranscriptionModel] {
switch self {
case .groq:
return [
TranscriptionModel(
id: "whisper-large-v3",
displayName: "Whisper Large V3",
description: "Default. Highest accuracy, best for detailed transcription."
),
TranscriptionModel(
id: "whisper-large-v3-turbo",
displayName: "Whisper Large V3 Turbo",
description: "Faster and cheaper. Slightly less accurate but great for most use cases."
),
]
case .openai:
return [
TranscriptionModel(
id: "whisper-1",
displayName: "Whisper 1",
description: "Default. OpenAI's general-purpose speech recognition model."
),
TranscriptionModel(
id: "gpt-4o-transcribe",
displayName: "GPT-4o Transcribe",
description: "Higher accuracy using GPT-4o. Better at handling accents and noisy audio."
),
TranscriptionModel(
id: "gpt-4o-mini-transcribe",
displayName: "GPT-4o Mini Transcribe",
description: "Cheaper GPT-4o variant. Good balance of cost and quality."
),
]
case .togetherAI:
return [
TranscriptionModel(
id: "openai/whisper-large-v3",
displayName: "Whisper Large V3",
description: "Default. High accuracy Whisper model hosted on Together AI."
),
TranscriptionModel(
id: "openai/whisper-large-v3-turbo",
displayName: "Whisper Large V3 Turbo",
description: "Faster variant with slightly reduced accuracy."
),
TranscriptionModel(
id: "deepgram/deepgram-nova-3",
displayName: "Deepgram Nova 3",
description: "Deepgram's latest model. Optimized for real-time, low-latency transcription."
),
TranscriptionModel(
id: "nvidia/parakeet-tdt-0.6b-v3",
displayName: "NVIDIA Parakeet",
description: "Lightweight 0.6B parameter model. Very fast with good English accuracy."
),
]
case .fireworks:
return [
TranscriptionModel(
id: "whisper-v3",
displayName: "Whisper V3",
description: "Default. Full Whisper V3 model for highest accuracy."
),
TranscriptionModel(
id: "whisper-v3-turbo",
displayName: "Whisper V3 Turbo",
description: "Faster variant. Good for shorter recordings where speed matters."
),
]
case .nvidia:
return [
TranscriptionModel(
id: "openai/whisper-large-v3",
displayName: "Whisper Large V3",
description: "Default. Whisper Large V3 running on NVIDIA GPU infrastructure."
),
]
case .custom:
return [
TranscriptionModel(
id: "whisper-large-v3",
displayName: "Whisper Large V3",
description: "Default. Common model ID for OpenAI-compatible endpoints."
),
TranscriptionModel(
id: "whisper-large-v3-turbo",
displayName: "Whisper Large V3 Turbo",
description: "Faster variant. Common model ID for OpenAI-compatible endpoints."
),
]
}
}

var availableChatModels: [ChatModel] {
switch self {
case .groq:
return [
ChatModel(
id: "meta-llama/llama-4-scout-17b-16e-instruct",
displayName: "Llama 4 Scout 17B",
description: "Default. Fast and capable model with vision support for context analysis."
),
]
case .openai:
return [
ChatModel(
id: "gpt-4o-mini",
displayName: "GPT-4o Mini",
description: "Default. Fast and affordable. Great for post-processing transcriptions."
),
ChatModel(
id: "gpt-4o",
displayName: "GPT-4o",
description: "Most capable. Better at complex context but slower and more expensive."
),
]
case .togetherAI:
return [
ChatModel(
id: "meta-llama/Llama-3.3-70B-Instruct-Turbo",
displayName: "Llama 3.3 70B Turbo",
description: "Default. High quality with fast inference. Best for accurate post-processing."
),
ChatModel(
id: "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
displayName: "Llama 3.1 8B Turbo",
description: "Smaller and faster. Good enough for simple dictation cleanup."
),
]
case .fireworks:
return [
ChatModel(
id: "accounts/fireworks/models/llama-v3p3-70b-instruct",
displayName: "Llama 3.3 70B",
description: "Default. High quality Llama model on Fireworks infrastructure."
),
]
case .nvidia:
return [
ChatModel(
id: "meta/llama-3.1-8b-instruct",
displayName: "Llama 3.1 8B",
description: "Default. Lightweight model running on NVIDIA GPU infrastructure."
),
]
case .custom:
return [
ChatModel(
id: "meta-llama/llama-4-scout-17b-16e-instruct",
displayName: "Llama 4 Scout 17B",
description: "Default. Common model ID for OpenAI-compatible endpoints."
),
]
}
}

var defaultTranscriptionModel: TranscriptionModel {
availableTranscriptionModels.first!
}

var defaultChatModel: ChatModel {
availableChatModels.first!
}

var apiKeyStorageKey: String {
switch self {
case .groq: return "groq_api_key"
case .openai: return "openai_api_key"
case .togetherAI: return "together_api_key"
case .fireworks: return "fireworks_api_key"
case .nvidia: return "nvidia_api_key"
case .custom: return "custom_api_key"
}
}

var apiKeyPlaceholder: String {
switch self {
case .groq: return "Paste your Groq API key"
case .openai: return "Paste your OpenAI API key"
case .togetherAI: return "Paste your Together AI API key"
case .fireworks: return "Paste your Fireworks API key"
case .nvidia: return "Paste your NVIDIA API key"
case .custom: return "Paste your API key"
}
}

var keyInstructionURL: URL? {
switch self {
case .groq: return URL(string: "https://console.groq.com/keys")
case .openai: return URL(string: "https://platform.openai.com/api-keys")
case .togetherAI: return URL(string: "https://api.together.ai/settings/api-keys")
case .fireworks: return URL(string: "https://fireworks.ai/account/api-keys")
case .nvidia: return URL(string: "https://build.nvidia.com/")
case .custom: return nil
}
}

var keyInstructionDisplayURL: String {
switch self {
case .groq: return "console.groq.com/keys"
case .openai: return "platform.openai.com/api-keys"
case .togetherAI: return "api.together.ai/settings/api-keys"
case .fireworks: return "fireworks.ai/account/api-keys"
case .nvidia: return "build.nvidia.com"
case .custom: return ""
}
}
}

// MARK: - Model Types

struct TranscriptionModel: Identifiable, Codable, Equatable, Hashable {
let id: String
let displayName: String
var description: String = ""

/// Sentinel value representing a user-entered custom model ID
static let customPlaceholder = TranscriptionModel(
id: "__custom__",
displayName: "Custom",
description: "Enter your own model ID."
)

var isCustom: Bool { id == "__custom__" }
}

struct ChatModel: Identifiable, Codable, Equatable, Hashable {
let id: String
let displayName: String
var description: String = ""

/// Sentinel value representing a user-entered custom model ID
static let customPlaceholder = ChatModel(
id: "__custom__",
displayName: "Custom",
description: "Enter your own model ID."
)

var isCustom: Bool { id == "__custom__" }
}
8 changes: 5 additions & 3 deletions Sources/AppContextService.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -31,15 +31,17 @@ Return only two sentences, no labels, no markdown, no extra commentary.
private let apiKey: String
private let baseURL: String
private let customContextPrompt: String
private let fallbackTextModel = "meta-llama/llama-4-scout-17b-16e-instruct"
private let visionModel = "meta-llama/llama-4-scout-17b-16e-instruct"
private let fallbackTextModel: String
private let visionModel: String
private let maxScreenshotDataURILength = 500_000
private let screenshotCompressionPrimary = 0.5
private let screenshotMaxDimension: CGFloat = 1024

init(apiKey: String, baseURL: String = "https://api.groq.com/openai/v1", customContextPrompt: String = "") {
init(apiKey: String, baseURL: String = "https://api.groq.com/openai/v1", chatModel: String = "meta-llama/llama-4-scout-17b-16e-instruct", visionModel: String = "meta-llama/llama-4-scout-17b-16e-instruct", customContextPrompt: String = "") {
self.apiKey = apiKey
self.baseURL = baseURL
self.fallbackTextModel = chatModel
self.visionModel = visionModel
self.customContextPrompt = customContextPrompt
}

Expand Down
Loading