Show token bar at bottom of chat input (#33)

src/components/chat/ChatInput.tsx

@@ -15,6 +15,7 @@ import {
   Database,
   ChevronsUpDown,
   ChevronsDownUp,
+  BarChart2,
 } from "lucide-react";
 import type React from "react";
 import { useCallback, useEffect, useRef, useState } from "react";
@@ -22,7 +23,7 @@ import { ModelPicker } from "@/components/ModelPicker";
 import { useSettings } from "@/hooks/useSettings";
 import { IpcClient } from "@/ipc/ipc_client";
 import { chatInputValueAtom, chatMessagesAtom } from "@/atoms/chatAtoms";
-import { useAtom, useSetAtom } from "jotai";
+import { atom, useAtom, useSetAtom } from "jotai";
 import { useStreamChat } from "@/hooks/useStreamChat";
 import { useChats } from "@/hooks/useChats";
 import { selectedAppIdAtom } from "@/atoms/appAtoms";
@@ -46,6 +47,9 @@ import { useRunApp } from "@/hooks/useRunApp";
 import { AutoApproveSwitch } from "../AutoApproveSwitch";
 import { usePostHog } from "posthog-js/react";
 import { CodeHighlight } from "./CodeHighlight";
+import { TokenBar } from "./TokenBar";
+
+const showTokenBarAtom = atom(false);
 
 export function ChatInput({ chatId }: { chatId?: number }) {
   const posthog = usePostHog();
@@ -60,6 +64,7 @@ export function ChatInput({ chatId }: { chatId?: number }) {
   const [isRejecting, setIsRejecting] = useState(false); // State for rejecting
   const [messages, setMessages] = useAtom<Message[]>(chatMessagesAtom);
   const setIsPreviewOpen = useSetAtom(isPreviewOpenAtom);
+  const [showTokenBar, setShowTokenBar] = useAtom(showTokenBarAtom);
 
   const { refreshAppIframe } = useRunApp();
 
@@ -274,14 +279,26 @@ export function ChatInput({ chatId }: { chatId?: number }) {
               </button>
             )}
           </div>
-          <div className="px-2 pb-2">
-            <ModelPicker
-              selectedModel={settings.selectedModel}
-              onModelSelect={(model) =>
-                updateSettings({ selectedModel: model })
-              }
-            />
+          <div className="pl-2 pr-1 flex items-center justify-between">
+            <div className="pb-2">
+              <ModelPicker
+                selectedModel={settings.selectedModel}
+                onModelSelect={(model) =>
+                  updateSettings({ selectedModel: model })
+                }
+              />
+            </div>
+            <button
+              onClick={() => setShowTokenBar(!showTokenBar)}
+              className="flex items-center px-2 py-1 text-xs text-muted-foreground hover:bg-muted rounded"
+              title={showTokenBar ? "Hide token usage" : "Show token usage"}
+            >
+              <BarChart2 size={14} className="mr-1" />
+              {showTokenBar ? "Hide tokens" : "Tokens"}
+            </button>
           </div>
+          {/* TokenBar is only displayed when showTokenBar is true */}
+          {showTokenBar && <TokenBar chatId={chatId} />}
         </div>
       </div>
     </>

src/components/chat/TokenBar.tsx

@@ -0,0 +1,130 @@
+import React, { useEffect, useState } from "react";
+import {
+  Tooltip,
+  TooltipContent,
+  TooltipProvider,
+  TooltipTrigger,
+} from "@/components/ui/tooltip";
+import { useCountTokens } from "@/hooks/useCountTokens";
+import { MessageSquare, Code, Bot, AlignLeft } from "lucide-react";
+import { chatInputValueAtom } from "@/atoms/chatAtoms";
+import { useAtom } from "jotai";
+import { useSettings } from "@/hooks/useSettings";
+
+interface TokenBarProps {
+  chatId?: number;
+}
+
+export function TokenBar({ chatId }: TokenBarProps) {
+  const [inputValue] = useAtom(chatInputValueAtom);
+  const { countTokens, result } = useCountTokens();
+  const [error, setError] = useState<string | null>(null);
+  const { settings } = useSettings();
+  useEffect(() => {
+    if (!chatId) return;
+    // Mark this as used, we need to re-trigger token count
+    // when selected model changes.
+    void settings?.selectedModel;
+
+    const debounceTimer = setTimeout(() => {
+      countTokens(chatId, inputValue).catch((err) => {
+        setError("Failed to count tokens");
+        console.error("Token counting error:", err);
+      });
+    }, 500);
+
+    return () => clearTimeout(debounceTimer);
+  }, [chatId, inputValue, countTokens, settings?.selectedModel]);
+
+  if (!chatId || !result) {
+    return null;
+  }
+
+  const {
+    totalTokens,
+    messageHistoryTokens,
+    codebaseTokens,
+    systemPromptTokens,
+    inputTokens,
+    contextWindow,
+  } = result;
+
+  const percentUsed = Math.min((totalTokens / contextWindow) * 100, 100);
+
+  // Calculate widths for each token type
+  const messageHistoryPercent = (messageHistoryTokens / contextWindow) * 100;
+  const codebasePercent = (codebaseTokens / contextWindow) * 100;
+  const systemPromptPercent = (systemPromptTokens / contextWindow) * 100;
+  const inputPercent = (inputTokens / contextWindow) * 100;
+
+  return (
+    <div className="px-4 pb-2 text-xs">
+      <TooltipProvider>
+        <Tooltip>
+          <TooltipTrigger asChild>
+            <div className="w-full">
+              <div className="flex justify-between mb-1 text-xs text-muted-foreground">
+                <span>Tokens: {totalTokens.toLocaleString()}</span>
+                <span>
+                  {Math.round(percentUsed)}% of{" "}
+                  {(contextWindow / 1000).toFixed(0)}K
+                </span>
+              </div>
+              <div className="w-full h-1.5 bg-muted rounded-full overflow-hidden flex">
+                {/* Message history tokens */}
+                <div
+                  className="h-full bg-blue-400"
+                  style={{ width: `${messageHistoryPercent}%` }}
+                />
+                {/* Codebase tokens */}
+                <div
+                  className="h-full bg-green-400"
+                  style={{ width: `${codebasePercent}%` }}
+                />
+                {/* System prompt tokens */}
+                <div
+                  className="h-full bg-purple-400"
+                  style={{ width: `${systemPromptPercent}%` }}
+                />
+                {/* Input tokens */}
+                <div
+                  className="h-full bg-yellow-400"
+                  style={{ width: `${inputPercent}%` }}
+                />
+              </div>
+            </div>
+          </TooltipTrigger>
+          <TooltipContent side="top" className="w-64 p-2">
+            <div className="space-y-1">
+              <div className="font-medium">Token Usage Breakdown</div>
+              <div className="grid grid-cols-[20px_1fr_auto] gap-x-2 items-center">
+                <MessageSquare size={12} className="text-blue-500" />
+                <span>Message History</span>
+                <span>{messageHistoryTokens.toLocaleString()}</span>
+
+                <Code size={12} className="text-green-500" />
+                <span>Codebase</span>
+                <span>{codebaseTokens.toLocaleString()}</span>
+
+                <Bot size={12} className="text-purple-500" />
+                <span>System Prompt</span>
+                <span>{systemPromptTokens.toLocaleString()}</span>
+
+                <AlignLeft size={12} className="text-yellow-500" />
+                <span>Current Input</span>
+                <span>{inputTokens.toLocaleString()}</span>
+              </div>
+              <div className="pt-1 border-t border-border">
+                <div className="flex justify-between font-medium">
+                  <span>Total</span>
+                  <span>{totalTokens.toLocaleString()}</span>
+                </div>
+              </div>
+            </div>
+          </TooltipContent>
+        </Tooltip>
+      </TooltipProvider>
+      {error && <div className="text-red-500 text-xs mt-1">{error}</div>}
+    </div>
+  );
+}

src/constants/models.ts

@@ -5,6 +5,7 @@ export interface ModelOption {
   description: string;
   tag?: string;
   maxOutputTokens?: number;
+  contextWindow?: number;
 }
 
 type RegularModelProvider = Exclude<ModelProvider, "ollama">;
@@ -16,6 +17,7 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       displayName: "GPT 4.1",
       description: "OpenAI's flagship model",
       maxOutputTokens: 32_768,
+      contextWindow: 1_047_576,
     },
     // https://platform.openai.com/docs/models/gpt-4.1-mini
     {
@@ -23,6 +25,7 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       displayName: "GPT 4.1 Mini",
       description: "OpenAI's lightweight, but intelligent model",
       maxOutputTokens: 32_768,
+      contextWindow: 1_047_576,
     },
     // https://platform.openai.com/docs/models/o3-mini
     {
@@ -30,6 +33,7 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       displayName: "o3 mini",
       description: "Reasoning model",
       maxOutputTokens: 100_000,
+      contextWindow: 200_000,
     },
   ],
   // https://docs.anthropic.com/en/docs/about-claude/models/all-models#model-comparison-table
@@ -39,6 +43,7 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       displayName: "Claude 3.7 Sonnet",
       description: "Excellent coder",
       maxOutputTokens: 64_000,
+      contextWindow: 200_000,
     },
   ],
   google: [
@@ -49,6 +54,8 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       description: "Experimental version of Google's Gemini 2.5 Pro model",
       tag: "Recommended",
       maxOutputTokens: 65_536,
+      // Gemini context window = input token + output token
+      contextWindow: 1_048_576,
     },
   ],
   openrouter: [
@@ -58,6 +65,7 @@ export const MODEL_OPTIONS: Record<RegularModelProvider, ModelOption[]> = {
       displayName: "DeepSeek v3 (free)",
       description: "Use for free (data may be used for training)",
       maxOutputTokens: 32_000,
+      contextWindow: 128_000,
     },
   ],
   auto: [

src/hooks/useCountTokens.ts

@@ -0,0 +1,43 @@
+import { useCallback } from "react";
+import { atom, useAtom } from "jotai";
+import { IpcClient } from "@/ipc/ipc_client";
+import type { TokenCountResult } from "@/ipc/ipc_types";
+
+// Create atoms to store the token count state
+export const tokenCountResultAtom = atom<TokenCountResult | null>(null);
+export const tokenCountLoadingAtom = atom<boolean>(false);
+export const tokenCountErrorAtom = atom<Error | null>(null);
+
+export function useCountTokens() {
+  const [result, setResult] = useAtom(tokenCountResultAtom);
+  const [loading, setLoading] = useAtom(tokenCountLoadingAtom);
+  const [error, setError] = useAtom(tokenCountErrorAtom);
+
+  const countTokens = useCallback(
+    async (chatId: number, input: string) => {
+      setLoading(true);
+      setError(null);
+
+      try {
+        const ipcClient = IpcClient.getInstance();
+        const tokenResult = await ipcClient.countTokens({ chatId, input });
+        setResult(tokenResult);
+        return tokenResult;
+      } catch (error) {
+        console.error("Error counting tokens:", error);
+        setError(error instanceof Error ? error : new Error(String(error)));
+        throw error;
+      } finally {
+        setLoading(false);
+      }
+    },
+    [setLoading, setError, setResult]
+  );
+
+  return {
+    countTokens,
+    result,
+    loading,
+    error,
+  };
+}

src/hooks/useStreamChat.ts

@@ -18,6 +18,7 @@ import { showError } from "@/lib/toast";
 import { useProposal } from "./useProposal";
 import { useSearch } from "@tanstack/react-router";
 import { useRunApp } from "./useRunApp";
+import { useCountTokens } from "./useCountTokens";
 
 export function getRandomNumberId() {
   return Math.floor(Math.random() * 1_000_000_000_000_000);
@@ -36,6 +37,8 @@ export function useStreamChat({
   const setStreamCount = useSetAtom(chatStreamCountAtom);
   const { refreshVersions } = useLoadVersions(selectedAppId);
   const { refreshAppIframe } = useRunApp();
+  const { countTokens } = useCountTokens();
+
   let chatId: number | undefined;
 
   if (hasChatId) {
@@ -111,6 +114,7 @@ export function useStreamChat({
             refreshChats();
             refreshApp();
             refreshVersions();
+            countTokens(chatId, "");
           },
           onError: (errorMessage: string) => {
             console.error(`[CHAT] Stream error for ${chatId}:`, errorMessage);
@@ -121,6 +125,7 @@ export function useStreamChat({
             refreshChats();
             refreshApp();
             refreshVersions();
+            countTokens(chatId, "");
           },
         });
       } catch (error) {

src/ipc/handlers/token_count_handlers.ts

@@ -0,0 +1,127 @@
+import { ipcMain } from "electron";
+import { db } from "../../db";
+import { chats, messages } from "../../db/schema";
+import { eq } from "drizzle-orm";
+import { SYSTEM_PROMPT } from "../../prompts/system_prompt";
+import {
+  SUPABASE_AVAILABLE_SYSTEM_PROMPT,
+  SUPABASE_NOT_AVAILABLE_SYSTEM_PROMPT,
+} from "../../prompts/supabase_prompt";
+import { getDyadAppPath } from "../../paths/paths";
+import log from "electron-log";
+import { extractCodebase } from "../../utils/codebase";
+import { getSupabaseContext } from "../../supabase_admin/supabase_context";
+import { readSettings } from "../../main/settings";
+import { MODEL_OPTIONS } from "../../constants/models";
+import { TokenCountParams } from "../ipc_types";
+import { TokenCountResult } from "../ipc_types";
+
+const logger = log.scope("token_count_handlers");
+
+// Estimate tokens (4 characters per token)
+const estimateTokens = (text: string): number => {
+  return Math.ceil(text.length / 4);
+};
+
+export function registerTokenCountHandlers() {
+  ipcMain.handle(
+    "chat:count-tokens",
+    async (event, req: TokenCountParams): Promise<TokenCountResult> => {
+      try {
+        // Get the chat with messages
+        const chat = await db.query.chats.findFirst({
+          where: eq(chats.id, req.chatId),
+          with: {
+            messages: {
+              orderBy: (messages, { asc }) => [asc(messages.createdAt)],
+            },
+            app: true,
+          },
+        });
+
+        if (!chat) {
+          throw new Error(`Chat not found: ${req.chatId}`);
+        }
+
+        // Prepare message history for token counting
+        const messageHistory = chat.messages
+          .map((message) => message.content)
+          .join("");
+        const messageHistoryTokens = estimateTokens(messageHistory);
+
+        // Count input tokens
+        const inputTokens = estimateTokens(req.input);
+
+        // Count system prompt tokens
+        let systemPrompt = SYSTEM_PROMPT;
+        let supabaseContext = "";
+
+        if (chat.app?.supabaseProjectId) {
+          systemPrompt += "\n\n" + SUPABASE_AVAILABLE_SYSTEM_PROMPT;
+          supabaseContext = await getSupabaseContext({
+            supabaseProjectId: chat.app.supabaseProjectId,
+          });
+        } else {
+          systemPrompt += "\n\n" + SUPABASE_NOT_AVAILABLE_SYSTEM_PROMPT;
+        }
+
+        const systemPromptTokens = estimateTokens(
+          systemPrompt + supabaseContext
+        );
+
+        // Extract codebase information if app is associated with the chat
+        let codebaseInfo = "";
+        let codebaseTokens = 0;
+
+        if (chat.app) {
+          const appPath = getDyadAppPath(chat.app.path);
+          try {
+            codebaseInfo = await extractCodebase(appPath);
+            codebaseTokens = estimateTokens(codebaseInfo);
+            logger.log(
+              `Extracted codebase information from ${appPath}, tokens: ${codebaseTokens}`
+            );
+          } catch (error) {
+            logger.error("Error extracting codebase:", error);
+          }
+        }
+
+        // Calculate total tokens
+        const totalTokens =
+          messageHistoryTokens +
+          inputTokens +
+          systemPromptTokens +
+          codebaseTokens;
+
+        return {
+          totalTokens,
+          messageHistoryTokens,
+          codebaseTokens,
+          inputTokens,
+          systemPromptTokens,
+          contextWindow: getContextWindow(),
+        };
+      } catch (error) {
+        logger.error("Error counting tokens:", error);
+        throw error;
+      }
+    }
+  );
+}
+
+const DEFAULT_CONTEXT_WINDOW = 128_000;
+
+function getContextWindow() {
+  const settings = readSettings();
+  const model = settings.selectedModel;
+  if (!MODEL_OPTIONS[model.provider as keyof typeof MODEL_OPTIONS]) {
+    logger.warn(
+      `Model provider ${model.provider} not found in MODEL_OPTIONS. Using default max tokens.`
+    );
+    return DEFAULT_CONTEXT_WINDOW;
+  }
+  const modelOption = MODEL_OPTIONS[
+    model.provider as keyof typeof MODEL_OPTIONS
+  ].find((m) => m.name === model.name);
+  return modelOption?.contextWindow || DEFAULT_CONTEXT_WINDOW;
+}

src/ipc/ipc_client.ts

@@ -19,6 +19,8 @@ import type {
   SystemDebugInfo,
   LocalModel,
   LocalModelListResponse,
+  TokenCountParams,
+  TokenCountResult,
 } from "./ipc_types";
 import type { CodeProposal, ProposalResult } from "@/lib/schemas";
 import { showError } from "@/lib/toast";
@@ -747,4 +749,17 @@ export class IpcClient {
       this.ipcRenderer.removeListener("deep-link-received", listener);
     };
   }
+
+  // Count tokens for a chat and input
+  public async countTokens(
+    params: TokenCountParams
+  ): Promise<TokenCountResult> {
+    try {
+      const result = await this.ipcRenderer.invoke("chat:count-tokens", params);
+      return result as TokenCountResult;
+    } catch (error) {
+      showError(error);
+      throw error;
+    }
+  }
 }

src/ipc/ipc_host.ts

@@ -10,6 +10,7 @@ import { registerProposalHandlers } from "./handlers/proposal_handlers";
 import { registerDebugHandlers } from "./handlers/debug_handlers";
 import { registerSupabaseHandlers } from "./handlers/supabase_handlers";
 import { registerLocalModelHandlers } from "./handlers/local_model_handlers";
+import { registerTokenCountHandlers } from "./handlers/token_count_handlers";
 
 export function registerIpcHandlers() {
   // Register all IPC handlers by category
@@ -25,4 +26,5 @@ export function registerIpcHandlers() {
   registerDebugHandlers();
   registerSupabaseHandlers();
   registerLocalModelHandlers();
+  registerTokenCountHandlers();
 }

src/ipc/ipc_types.ts

@@ -101,3 +101,16 @@ export type LocalModelListResponse = {
   models: LocalModel[];
   error: string | null;
 };
+
+export interface TokenCountParams {
+  chatId: number;
+  input: string;
+}
+export interface TokenCountResult {
+  totalTokens: number;
+  messageHistoryTokens: number;
+  codebaseTokens: number;
+  inputTokens: number;
+  systemPromptTokens: number;
+  contextWindow: number;
+}

src/preload.ts

@@ -9,6 +9,7 @@ const validInvokeChannels = [
   "chat:message",
   "chat:cancel",
   "chat:stream",
+  "chat:count-tokens",
   "create-chat",
   "create-app",
   "get-chat",