MindWorkAI/AI-Studio
MindWorkAI/AI-Studio
1
0
Fork 0
mirror of https://github.com/MindWorkAI/AI-Studio.git synced 2026-02-12 07:01:37 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Added the OpenRouter provider (#595)

Browse Source
This commit is contained in:
Thorsten Sommer 2025-12-15 19:31:51 +01:00 committed by GitHub
parent cfad42ddf4
commit a39c6c330f
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
10 changed files with 495 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
README.md
View File

@ -32,7 +32,7 @@ Since November 2024: Work on RAG (integration of your data and files) has begun.
- [x] ~~App: Implement dialog for checking & handling [pandoc](https://pandoc.org/) installation ([PR #393](https://github.com/MindWorkAI/AI-Studio/pull/393), [PR #487](https://github.com/MindWorkAI/AI-Studio/pull/487))~~
- [ ] App: Implement external embedding providers
- [ ] App: Implement the process to vectorize one local file using embeddings
- [ ] Runtime: Integration of the vector database [LanceDB](https://github.com/lancedb/lancedb)
- [ ] Runtime: Integration of the vector database [Qdrant](https://github.com/qdrant/qdrant)
- [ ] App: Implement the continuous process of vectorizing data
- [x] ~~App: Define a common retrieval context interface for the integration of RAG processes in chats (PR [#281](https://github.com/MindWorkAI/AI-Studio/pull/281), [#284](https://github.com/MindWorkAI/AI-Studio/pull/284), [#286](https://github.com/MindWorkAI/AI-Studio/pull/286), [#287](https://github.com/MindWorkAI/AI-Studio/pull/287))~~
- [x] ~~App: Define a common augmentation interface for the integration of RAG processes in chats (PR [#288](https://github.com/MindWorkAI/AI-Studio/pull/288), [#289](https://github.com/MindWorkAI/AI-Studio/pull/289))~~
@ -78,7 +78,7 @@ Since March 2025: We have started developing the plugin system. There will be la
Features we have recently released
</h3>
</summary>
- v0.9.55: Added support for newer models like Mistral 3 & GPT 5.2, OpenRouter as LLM and embedding provider, and the possibility to use file attachments in chats.
- v0.9.51: Added support for [Perplexity](https://www.perplexity.ai/); citations added so that LLMs can provide source references (e.g., some OpenAI models, Perplexity); added support for OpenAI's Responses API so that all text LLMs from OpenAI now work in MindWork AI Studio, including Deep Research models; web searches are now possible (some OpenAI models, Perplexity).
- v0.9.50: Added support for self-hosted LLMs using [vLLM](https://blog.vllm.ai/2023/06/20/vllm.html).
- v0.9.46: Released our plugin system, a German language plugin, early support for enterprise environments, and configuration plugins. Additionally, we added the Pandoc integration for future data processing and file generation.
@ -114,6 +114,7 @@ MindWork AI Studio is a free desktop app for macOS, Windows, and Linux. It provi
- [xAI](https://x.ai/) (Grok)
- [DeepSeek](https://www.deepseek.com/en)
- [Alibaba Cloud](https://www.alibabacloud.com) (Qwen)
- [OpenRouter](https://openrouter.ai/)
- [Hugging Face](https://huggingface.co/) using their [inference providers](https://huggingface.co/docs/inference-providers/index) such as Cerebras, Nebius, Sambanova, Novita, Hyperbolic, Together AI, Fireworks, Hugging Face
- Self-hosted models using [llama.cpp](https://github.com/ggerganov/llama.cpp), [ollama](https://github.com/ollama/ollama), [LM Studio](https://lmstudio.ai/), and [vLLM](https://github.com/vllm-project/vllm)
- [Groq](https://groq.com/)

2
app/MindWork AI Studio/Pages/Home.razor.cs
View File

@ -31,7 +31,7 @@ public partial class Home : MSGComponentBase
{
this.itemsAdvantages = [
new(this.T("Free of charge"), this.T("The app is free to use, both for personal and commercial purposes.")),
new(this.T("Independence"), this.T("You are not tied to any single provider. Instead, you might choose the provider that best suits your needs. Right now, we support OpenAI (GPT5, o1, etc.), Perplexity, Mistral, Anthropic (Claude), Google Gemini, xAI (Grok), DeepSeek, Alibaba Cloud (Qwen), Hugging Face, and self-hosted models using vLLM, llama.cpp, ollama, LM Studio, Groq, or Fireworks. For scientists and employees of research institutions, we also support Helmholtz and GWDG AI services. These are available through federated logins like eduGAIN to all 18 Helmholtz Centers, the Max Planck Society, most German, and many international universities.")),
new(this.T("Independence"), this.T("You are not tied to any single provider. Instead, you might choose the provider that best suits your needs. Right now, we support OpenAI (GPT5, o1, etc.), Perplexity, Mistral, Anthropic (Claude), Google Gemini, xAI (Grok), DeepSeek, Alibaba Cloud (Qwen), OpenRouter, Hugging Face, and self-hosted models using vLLM, llama.cpp, ollama, LM Studio, Groq, or Fireworks. For scientists and employees of research institutions, we also support Helmholtz and GWDG AI services. These are available through federated logins like eduGAIN to all 18 Helmholtz Centers, the Max Planck Society, most German, and many international universities.")),
new(this.T("Assistants"), this.T("You just want to quickly translate a text? AI Studio has so-called assistants for such and other tasks. No prompting is necessary when working with these assistants.")),
new(this.T("Unrestricted usage"), this.T("Unlike services like ChatGPT, which impose limits after intensive use, MindWork AI Studio offers unlimited usage through the providers API.")),
new(this.T("Cost-effective"), this.T("You only pay for what you use, which can be cheaper than monthly subscription services like ChatGPT Plus, especially if used infrequently. But beware, here be dragons: For extremely intensive usage, the API costs can be significantly higher. Unfortunately, providers currently do not offer a way to display current costs in the app. Therefore, check your account with the respective provider to see how your costs are developing. When available, use prepaid and set a cost limit.")),

1
app/MindWork AI Studio/Provider/LLMProviders.cs
View File

@ -15,6 +15,7 @@ public enum LLMProviders
DEEP_SEEK = 11,
ALIBABA_CLOUD = 12,
PERPLEXITY = 14,
OPEN_ROUTER = 15,
FIREWORKS = 5,
GROQ = 6,

11
app/MindWork AI Studio/Provider/LLMProvidersExtensions.cs
View File

@ -9,6 +9,7 @@ using AIStudio.Provider.Helmholtz;
using AIStudio.Provider.HuggingFace;
using AIStudio.Provider.Mistral;
using AIStudio.Provider.OpenAI;
using AIStudio.Provider.OpenRouter;
using AIStudio.Provider.Perplexity;
using AIStudio.Provider.SelfHosted;
using AIStudio.Provider.X;
@ -42,6 +43,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => "DeepSeek",
LLMProviders.ALIBABA_CLOUD => "Alibaba Cloud",
LLMProviders.PERPLEXITY => "Perplexity",
LLMProviders.OPEN_ROUTER => "OpenRouter",
LLMProviders.GROQ => "Groq",
LLMProviders.FIREWORKS => "Fireworks.ai",
@ -93,6 +95,8 @@ public static class LLMProvidersExtensions
LLMProviders.PERPLEXITY => Confidence.USA_NO_TRAINING.WithRegion("America, U.S.").WithSources("https://www.perplexity.ai/hub/legal/perplexity-api-terms-of-service").WithLevel(settingsManager.GetConfiguredConfidenceLevel(llmProvider)),
LLMProviders.OPEN_ROUTER => Confidence.USA_HUB.WithRegion("America, U.S.").WithSources("https://openrouter.ai/privacy", "https://openrouter.ai/terms").WithLevel(settingsManager.GetConfiguredConfidenceLevel(llmProvider)),
LLMProviders.SELF_HOSTED => Confidence.SELF_HOSTED.WithLevel(settingsManager.GetConfiguredConfidenceLevel(llmProvider)),
LLMProviders.HELMHOLTZ => Confidence.GDPR_NO_TRAINING.WithRegion("Europe, Germany").WithSources("https://helmholtz.cloud/services/?serviceID=d7d5c597-a2f6-4bd1-b71e-4d6499d98570").WithLevel(settingsManager.GetConfiguredConfidenceLevel(llmProvider)),
@ -128,6 +132,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => false,
LLMProviders.HUGGINGFACE => false,
LLMProviders.PERPLEXITY => false,
LLMProviders.OPEN_ROUTER => true,
//
// Self-hosted providers are treated as a special case anyway.
@ -171,6 +176,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => new ProviderDeepSeek { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
LLMProviders.ALIBABA_CLOUD => new ProviderAlibabaCloud { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
LLMProviders.PERPLEXITY => new ProviderPerplexity { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
LLMProviders.OPEN_ROUTER => new ProviderOpenRouter { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
LLMProviders.GROQ => new ProviderGroq { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
LLMProviders.FIREWORKS => new ProviderFireworks { InstanceName = instanceName, AdditionalJsonApiParameters = expertProviderApiParameter },
@ -201,6 +207,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => "https://platform.deepseek.com/sign_up",
LLMProviders.ALIBABA_CLOUD => "https://account.alibabacloud.com/register/intl_register.htm",
LLMProviders.PERPLEXITY => "https://www.perplexity.ai/account/api",
LLMProviders.OPEN_ROUTER => "https://openrouter.ai/keys",
LLMProviders.GROQ => "https://console.groq.com/",
LLMProviders.FIREWORKS => "https://fireworks.ai/login",
@ -224,6 +231,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => "https://platform.deepseek.com/usage",
LLMProviders.ALIBABA_CLOUD => "https://usercenter2-intl.aliyun.com/billing",
LLMProviders.PERPLEXITY => "https://www.perplexity.ai/account/api/",
LLMProviders.OPEN_ROUTER => "https://openrouter.ai/activity",
LLMProviders.HUGGINGFACE => "https://huggingface.co/settings/billing",
_ => string.Empty,
@ -241,6 +249,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => true,
LLMProviders.ALIBABA_CLOUD => true,
LLMProviders.PERPLEXITY => true,
LLMProviders.OPEN_ROUTER => true,
LLMProviders.HUGGINGFACE => true,
_ => false,
@ -288,6 +297,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => true,
LLMProviders.ALIBABA_CLOUD => true,
LLMProviders.PERPLEXITY => true,
LLMProviders.OPEN_ROUTER => true,
LLMProviders.GROQ => true,
LLMProviders.FIREWORKS => true,
@ -310,6 +320,7 @@ public static class LLMProvidersExtensions
LLMProviders.DEEP_SEEK => true,
LLMProviders.ALIBABA_CLOUD => true,
LLMProviders.PERPLEXITY => true,
LLMProviders.OPEN_ROUTER => true,
LLMProviders.GROQ => true,
LLMProviders.FIREWORKS => true,

8
app/MindWork AI Studio/Provider/OpenRouter/OpenRouterModel.cs Normal file
View File

@ -0,0 +1,8 @@
namespace AIStudio.Provider.OpenRouter;
/// <summary>
/// A data model for an OpenRouter model from the API.
/// </summary>
/// <param name="Id">The model's ID.</param>
/// <param name="Name">The model's human-readable display name.</param>
public readonly record struct OpenRouterModel(string Id, string? Name);

7
app/MindWork AI Studio/Provider/OpenRouter/OpenRouterModelsResponse.cs Normal file
View File

@ -0,0 +1,7 @@
namespace AIStudio.Provider.OpenRouter;
/// <summary>
/// A data model for the response from the OpenRouter models endpoint.
/// </summary>
/// <param name="Data">The list of models.</param>
public readonly record struct OpenRouterModelsResponse(IList<OpenRouterModel> Data);

202
app/MindWork AI Studio/Provider/OpenRouter/ProviderOpenRouter.cs Normal file
View File

@ -0,0 +1,202 @@
using System.Net.Http.Headers;
using System.Runtime.CompilerServices;
using System.Text;
using System.Text.Json;
using AIStudio.Chat;
using AIStudio.Provider.OpenAI;
using AIStudio.Settings;
namespace AIStudio.Provider.OpenRouter;
public sealed class ProviderOpenRouter() : BaseProvider("https://openrouter.ai/api/v1/", LOGGER)
{
private const string PROJECT_WEBSITE = "https://github.com/MindWorkAI/AI-Studio";
private const string PROJECT_NAME = "MindWork AI Studio";
private static readonly ILogger<ProviderOpenRouter> LOGGER = Program.LOGGER_FACTORY.CreateLogger<ProviderOpenRouter>();
#region Implementation of IProvider
/// <inheritdoc />
public override string Id => LLMProviders.OPEN_ROUTER.ToName();
/// <inheritdoc />
public override string InstanceName { get; set; } = "OpenRouter";
/// <inheritdoc />
public override async IAsyncEnumerable<ContentStreamChunk> StreamChatCompletion(Model chatModel, ChatThread chatThread, SettingsManager settingsManager, [EnumeratorCancellation] CancellationToken token = default)
{
// Get the API key:
var requestedSecret = await RUST_SERVICE.GetAPIKey(this);
if(!requestedSecret.Success)
yield break;
// Prepare the system prompt:
var systemPrompt = new Message
{
Role = "system",
Content = chatThread.PrepareSystemPrompt(settingsManager, chatThread),
};
// Parse the API parameters:
var apiParameters = this.ParseAdditionalApiParameters();
// Build the list of messages:
var messages = await chatThread.Blocks.BuildMessages(async n => new Message
{
Role = n.Role switch
{
ChatRole.USER => "user",
ChatRole.AI => "assistant",
ChatRole.AGENT => "assistant",
ChatRole.SYSTEM => "system",
_ => "user",
},
Content = n.Content switch
{
ContentText text => await text.PrepareContentForAI(),
_ => string.Empty,
}
});
// Prepare the OpenRouter HTTP chat request:
var openRouterChatRequest = JsonSerializer.Serialize(new ChatCompletionAPIRequest
{
Model = chatModel.Id,
// Build the messages:
// - First of all the system prompt
// - Then none-empty user and AI messages
Messages = [systemPrompt, ..messages],
// Right now, we only support streaming completions:
Stream = true,
AdditionalApiParameters = apiParameters
}, JSON_SERIALIZER_OPTIONS);
async Task<HttpRequestMessage> RequestBuilder()
{
// Build the HTTP post request:
var request = new HttpRequestMessage(HttpMethod.Post, "chat/completions");
// Set the authorization header:
request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", await requestedSecret.Secret.Decrypt(ENCRYPTION));
// Set custom headers for project identification:
request.Headers.Add("HTTP-Referer", PROJECT_WEBSITE);
request.Headers.Add("X-Title", PROJECT_NAME);
// Set the content:
request.Content = new StringContent(openRouterChatRequest, Encoding.UTF8, "application/json");
return request;
}
await foreach (var content in this.StreamChatCompletionInternal<ChatCompletionDeltaStreamLine, NoChatCompletionAnnotationStreamLine>("OpenRouter", RequestBuilder, token))
yield return content;
}
#pragma warning disable CS1998 // Async method lacks 'await' operators and will run synchronously
/// <inheritdoc />
public override async IAsyncEnumerable<ImageURL> StreamImageCompletion(Model imageModel, string promptPositive, string promptNegative = FilterOperator.String.Empty, ImageURL referenceImageURL = default, [EnumeratorCancellation] CancellationToken token = default)
{
yield break;
}
#pragma warning restore CS1998 // Async method lacks 'await' operators and will run synchronously
/// <inheritdoc />
public override Task<IEnumerable<Model>> GetTextModels(string? apiKeyProvisional = null, CancellationToken token = default)
{
return this.LoadModels(token, apiKeyProvisional);
}
/// <inheritdoc />
public override Task<IEnumerable<Model>> GetImageModels(string? apiKeyProvisional = null, CancellationToken token = default)
{
return Task.FromResult(Enumerable.Empty<Model>());
}
/// <inheritdoc />
public override Task<IEnumerable<Model>> GetEmbeddingModels(string? apiKeyProvisional = null, CancellationToken token = default)
{
return this.LoadEmbeddingModels(token, apiKeyProvisional);
}
#endregion
private async Task<IEnumerable<Model>> LoadModels(CancellationToken token, string? apiKeyProvisional = null)
{
var secretKey = apiKeyProvisional switch
{
not null => apiKeyProvisional,
_ => await RUST_SERVICE.GetAPIKey(this) switch
{
{ Success: true } result => await result.Secret.Decrypt(ENCRYPTION),
_ => null,
}
};
if (secretKey is null)
return [];
using var request = new HttpRequestMessage(HttpMethod.Get, "models");
request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", secretKey);
// Set custom headers for project identification:
request.Headers.Add("HTTP-Referer", PROJECT_WEBSITE);
request.Headers.Add("X-Title", PROJECT_NAME);
using var response = await this.httpClient.SendAsync(request, token);
if(!response.IsSuccessStatusCode)
return [];
var modelResponse = await response.Content.ReadFromJsonAsync<OpenRouterModelsResponse>(token);
// Filter out non-text models (image, audio, embedding models) and convert to Model
return modelResponse.Data
.Where(n =>
!n.Id.Contains("whisper", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("dall-e", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("tts", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("embedding", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("moderation", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("stable-diffusion", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("flux", StringComparison.OrdinalIgnoreCase) &&
!n.Id.Contains("midjourney", StringComparison.OrdinalIgnoreCase))
.Select(n => new Model(n.Id, n.Name));
}
private async Task<IEnumerable<Model>> LoadEmbeddingModels(CancellationToken token, string? apiKeyProvisional = null)
{
var secretKey = apiKeyProvisional switch
{
not null => apiKeyProvisional,
_ => await RUST_SERVICE.GetAPIKey(this) switch
{
{ Success: true } result => await result.Secret.Decrypt(ENCRYPTION),
_ => null,
}
};
if (secretKey is null)
return [];
using var request = new HttpRequestMessage(HttpMethod.Get, "embeddings/models");
request.Headers.Authorization = new AuthenticationHeaderValue("Bearer", secretKey);
// Set custom headers for project identification:
request.Headers.Add("HTTP-Referer", PROJECT_WEBSITE);
request.Headers.Add("X-Title", PROJECT_NAME);
using var response = await this.httpClient.SendAsync(request, token);
if(!response.IsSuccessStatusCode)
return [];
var modelResponse = await response.Content.ReadFromJsonAsync<OpenRouterModelsResponse>(token);
// Convert all embedding models to Model
return modelResponse.Data.Select(n => new Model(n.Id, n.Name));
}
}

249
app/MindWork AI Studio/Settings/ProviderExtensions.OpenRouter.cs Normal file
View File

@ -0,0 +1,249 @@
using AIStudio.Provider;
namespace AIStudio.Settings;
public static partial class ProviderExtensions
{
public static List<Capability> GetModelCapabilitiesOpenRouter(Model model)
{
var modelName = model.Id.ToLowerInvariant().AsSpan();
//
// OpenRouter model IDs follow the pattern: "provider/model-name"
// Examples:
// - openai/gpt-4o
// - anthropic/claude-3-5-sonnet
// - google/gemini-pro-1.5
// - meta-llama/llama-3.1-405b-instruct
//
// We need to detect capabilities based on both provider and model name.
//
//
// OpenAI models via OpenRouter:
//
if (modelName.IndexOf("openai/") is not -1)
{
// Reasoning models (o1, o3, o4 series)
if (modelName.IndexOf("/o1") is not -1 ||
modelName.IndexOf("/o3") is not -1 ||
modelName.IndexOf("/o4") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.ALWAYS_REASONING,
Capability.CHAT_COMPLETION_API,
];
// GPT-4o and GPT-5 series with multimodal
if (modelName.IndexOf("/gpt-4o") is not -1 ||
modelName.IndexOf("/gpt-5") is not -1 ||
modelName.IndexOf("/chatgpt-4o") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// Standard GPT-4
if (modelName.IndexOf("/gpt-4") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// GPT-3.5
if (modelName.IndexOf("/gpt-3.5") is not -1 ||
modelName.IndexOf("/gpt-3") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
//
// Anthropic models via OpenRouter:
//
if (modelName.IndexOf("anthropic/") is not -1)
{
// Claude 3.5 and newer with vision
if (modelName.IndexOf("/claude-3.5") is not -1 ||
modelName.IndexOf("/claude-3-5") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// Claude 3 Opus/Sonnet with vision
if (modelName.IndexOf("/claude-3-opus") is not -1 ||
modelName.IndexOf("/claude-3-sonnet") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// Other Claude 3 models
if (modelName.IndexOf("/claude-3") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
//
// Google models via OpenRouter:
//
if (modelName.IndexOf("google/") is not -1)
{
// Gemini models with multimodal
if (modelName.IndexOf("/gemini") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
}
//
// xAI Grok models via OpenRouter:
//
if (modelName.IndexOf("x-ai/") is not -1 || modelName.IndexOf("/grok") is not -1)
{
if (modelName.IndexOf("-vision") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
}
//
// DeepSeek models via OpenRouter:
//
if (modelName.IndexOf("/deepseek") is not -1)
{
if (modelName.IndexOf("-r1") is not -1 || modelName.IndexOf(" r1") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.ALWAYS_REASONING,
Capability.CHAT_COMPLETION_API,
];
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
//
// Mistral models via OpenRouter:
//
if (modelName.IndexOf("/mistral") is not -1 || modelName.IndexOf("/pixtral") is not -1)
{
if (modelName.IndexOf("/pixtral") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
}
//
// Meta Llama models via OpenRouter:
//
if (modelName.IndexOf("/llama") is not -1)
{
// Llama 4 with vision
if (modelName.IndexOf("/llama-4") is not -1 ||
modelName.IndexOf("/llama4") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// Vision models
if (modelName.IndexOf("-vision") is not -1)
return [
Capability.TEXT_INPUT, Capability.MULTIPLE_IMAGE_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
// Llama 3.1+ with function calling
if (modelName.IndexOf("/llama-3.") is not -1 ||
modelName.IndexOf("/llama3.") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.FUNCTION_CALLING,
Capability.CHAT_COMPLETION_API,
];
// Default Llama
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
//
// Qwen models via OpenRouter:
//
if (modelName.IndexOf("/qwen") is not -1 || modelName.IndexOf("/qwq") is not -1)
{
if (modelName.IndexOf("/qwq") is not -1)
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.ALWAYS_REASONING,
Capability.CHAT_COMPLETION_API,
];
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
//
// Default for unknown models:
// Assume basic text input/output with chat completion
//
return [
Capability.TEXT_INPUT,
Capability.TEXT_OUTPUT,
Capability.CHAT_COMPLETION_API,
];
}
}

1
app/MindWork AI Studio/Settings/ProviderExtensions.cs
View File

@ -14,6 +14,7 @@ public static partial class ProviderExtensions
LLMProviders.DEEP_SEEK => GetModelCapabilitiesDeepSeek(provider.Model),
LLMProviders.ALIBABA_CLOUD => GetModelCapabilitiesAlibaba(provider.Model),
LLMProviders.PERPLEXITY => GetModelCapabilitiesPerplexity(provider.Model),
LLMProviders.OPEN_ROUTER => GetModelCapabilitiesOpenRouter(provider.Model),
LLMProviders.GROQ => GetModelCapabilitiesOpenSource(provider.Model),
LLMProviders.FIREWORKS => GetModelCapabilitiesOpenSource(provider.Model),

1
app/MindWork AI Studio/wwwroot/changelog/v0.9.55.md
View File

@ -1,6 +1,7 @@
# v0.9.55, build 230 (2025-12-xx xx:xx UTC)
- Added support for newer Mistral models (Mistral 3, Voxtral, and Magistral).
- Added support for the new OpenAI model GPT 5.2.
- Added support for OpenRouter as LLM and embedding provider.
- Added a description field to local data sources (preview feature) so that the data selection agent has more information about which data each local source contains when selecting data sources.
- Added the ability to use file attachments in chat. This is the initial implementation of this feature. We will continue to develop this feature and refine it further based on user feedback. Many thanks to Sabrina `Sabrina-devops` for this wonderful contribution.
- Improved the document analysis assistant (in preview) by adding descriptions to the different sections.