Merge remote-tracking branch 'upstream/main'

app/lib/.server/llm/api-key.ts

@@ -35,6 +35,8 @@ export function getAPIKey(cloudflareEnv: Env, provider: string, userApiKeys?: Re
       return env.OPENAI_LIKE_API_KEY || cloudflareEnv.OPENAI_LIKE_API_KEY;
     case "xAI":
       return env.XAI_API_KEY || cloudflareEnv.XAI_API_KEY;
+    case "Cohere":
+      return env.COHERE_API_KEY;
     default:
       return "";
   }

app/lib/.server/llm/model.ts

@@ -7,6 +7,11 @@ import { createGoogleGenerativeAI } from '@ai-sdk/google';
 import { ollama } from 'ollama-ai-provider';
 import { createOpenRouter } from "@openrouter/ai-sdk-provider";
 import { createMistral } from '@ai-sdk/mistral';
+import { createCohere } from '@ai-sdk/cohere'
+
+export const DEFAULT_NUM_CTX = process.env.DEFAULT_NUM_CTX ? 
+  parseInt(process.env.DEFAULT_NUM_CTX, 10) : 
+  32768;
 
 export function getAnthropicModel(apiKey: string, model: string) {
   const anthropic = createAnthropic({
@@ -22,14 +27,16 @@ export function getOpenAILikeModel(baseURL: string, apiKey: string, model: strin
     baseURL,
     apiKey,
   });
-  // console.log('OpenAI client created:', !!openai);
-  const client = openai(model);
-  // console.log('OpenAI model client:', !!client);
-  return client;
-  // return {
-  //   model: client,
-  //   provider: 'OpenAILike'  // Correctly identifying the actual provider
-  // };
+
+  return openai(model);
+}
+
+export function getCohereAIModel(apiKey:string, model: string){
+  const cohere = createCohere({
+    apiKey,
+  });
+
+  return cohere(model);
 }
 
 export function getOpenAIModel(apiKey: string, model: string) {
@@ -76,7 +83,7 @@ export function getHuggingFaceModel(apiKey: string, model: string) {
 
 export function getOllamaModel(baseURL: string, model: string) {
   let Ollama = ollama(model, {
-    numCtx: 32768,
+    numCtx: DEFAULT_NUM_CTX,
   });
 
   Ollama.config.baseURL = `${baseURL}/api`;
@@ -150,6 +157,8 @@ export function getModel(provider: string, model: string, env: Env, apiKeys?: Re
       return getLMStudioModel(baseURL, model);
     case 'xAI':
       return getXAIModel(apiKey, model);
+    case 'Cohere':
+      return getCohereAIModel(apiKey, model);
     default:
       return getOllamaModel(baseURL, model);
   }

app/lib/.server/llm/stream-text.ts

@@ -58,7 +58,6 @@ function extractPropertiesFromMessage(message: Message): { model: string; provid
 
   return { model, provider, content: cleanedContent };
 }
-
 export function streamText(
   messages: Messages,
   env: Env,
@@ -68,8 +67,6 @@ export function streamText(
   let currentModel = DEFAULT_MODEL;
   let currentProvider = DEFAULT_PROVIDER;
 
-  // console.log('StreamText:', JSON.stringify(messages));
-
   const processedMessages = messages.map((message) => {
     if (message.role === 'user') {
       const { model, provider, content } = extractPropertiesFromMessage(message);
@@ -83,25 +80,19 @@ export function streamText(
       return { ...message, content };
     }
 
-    return message; // No changes for non-user messages
-  });
+    const modelDetails = MODEL_LIST.find((m) => m.name === currentModel);
 
-  // console.log('Message content:', messages[0].content);
-  // console.log('Extracted properties:', extractPropertiesFromMessage(messages[0]));
+    const dynamicMaxTokens =
+      modelDetails && modelDetails.maxTokenAllowed
+        ? modelDetails.maxTokenAllowed
+        : MAX_TOKENS;
 
-  const llmClient = getModel(currentProvider, currentModel, env, apiKeys);
-  // console.log('LLM Client:', llmClient);
-
-  const llmConfig = {
-    ...options,
-    model: llmClient, //getModel(currentProvider, currentModel, env, apiKeys),
-    provider: currentProvider,
-    system: getSystemPrompt(),
-    maxTokens: MAX_TOKENS,
-    messages: convertToCoreMessages(processedMessages),
-  };
-
-  // console.log('LLM Config:', llmConfig);
-
-  return _streamText(llmConfig);
-}
+    return _streamText({
+      model: getModel(currentProvider, currentModel, env, apiKeys),
+      system: getSystemPrompt(),
+      maxTokens: dynamicMaxTokens,
+      messages: convertToCoreMessages(processedMessages),
+      ...options,
+    });
+  }
+)}