Read the context window from four more model lists

app/MindWork AI Studio/Provider/OpenRouter/OpenRouterModel.cs

@@ -1,8 +1,16 @@
+using System.Text.Json.Serialization;
+
 namespace AIStudio.Provider.OpenRouter;
 
 /// <summary>
 /// A data model for an OpenRouter model from the API.
 /// </summary>
+/// <remarks>
+/// The window is the model's, not that of any one provider behind it. OpenRouter also states a
+/// window per provider it currently prefers, but it picks one per request, so a number taken from
+/// there would describe a choice nobody has made yet.
+/// </remarks>
 /// <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);
+/// <param name="ContextWindowTokens">How much the model reads and writes in one conversation, in tokens.</param>
+public readonly record struct OpenRouterModel(string Id, string? Name, [property: JsonPropertyName("context_length")] int? ContextWindowTokens);

app/MindWork AI Studio/Provider/OpenRouter/ProviderOpenRouter.cs

@@ -2,6 +2,7 @@ using System.Net.Http.Headers;
 using System.Runtime.CompilerServices;
 
 using AIStudio.Chat;
+using AIStudio.Models.Live;
 using AIStudio.Provider.OpenAI;
 using AIStudio.Settings;
 
@@ -125,9 +126,21 @@ public sealed class ProviderOpenRouter() : BaseProvider(LLMProviders.OPEN_ROUTER
                 request.Headers.Add("HTTP-Referer", PROJECT_WEBSITE);
                 request.Headers.Add("X-Title", PROJECT_NAME);
             },
+            listingFactory: modelResponse => modelResponse.Data.Select(n => ModelListing.For(n.Id, n.ContextWindowTokens)),
             token: token);
     }
 
+    /// <summary>
+    /// Loads the models OpenRouter offers for embedding, which live on a route of their own.
+    /// </summary>
+    /// <remarks>
+    /// Nothing is reported from here: this route answers with the embedding models alone, and what
+    /// is reported replaces everything an instance said before. The windows of the chat models
+    /// would go missing the moment somebody opens the embedding settings.
+    /// </remarks>
+    /// <param name="apiKeyProvisional">An API key which is not stored yet.</param>
+    /// <param name="token">The cancellation token to use.</param>
+    /// <returns>The embedding models.</returns>
     private Task<ModelLoadResult> LoadEmbeddingModels(string? apiKeyProvisional, CancellationToken token)
     {
         return this.LoadModelsResponse<OpenRouterModelsResponse>(