Search

                var providerMap = {
                  'groq': 'Groq',
                  'openai': 'OpenAI',
                  'anthropic': 'Anthropic',
                  'custom': 'Custom Provider'

  { id: 'groq/compound-mini', name: 'Compound Mini', desc: 'Faster, lighter version with web search and code execution.', logo: '/public/groq-logo-orange.png', provider: 'groq', web: true, tools: true },
  
  // Groq - OpenAI GPT-OSS Models
  { id: 'openai/gpt-oss-120b', name: 'GPT OSS 120B', desc: 'Flagship 120B parameter model with browser search and reasoning.', logo: '/public/groq-logo-orange.png', provider: 'groq', web: true, thinking: true, tools: true },
  { id: 'openai/gpt-oss-20b', name: 'GPT OSS 20B', desc: 'Fast 20B model with web search, thinking, and code execution.', logo: '/public/groq-logo-orange.png', provider: 'groq', web: true, thinking: true, tools: true },
  
  // Groq - Llama 3.x Production
  { id: 'meta-llama/llama-4-scout-17b-16e-instruct', name: 'Llama 4 Scout 17B', desc: 'Fast responses with image support (750 t/s).', logo: '/public/groq-logo-orange.png', provider: 'groq', image: true, preview: true },
  
  // OpenAI - GPT-5.2 Series (Latest)
  { id: 'gpt-5.2', name: 'GPT-5.2', desc: 'Most advanced model for coding and agentic tasks across industries.', logo: '/public/openai-logo.png', provider: 'openai', image: true, tools: true },
  { id: 'gpt-5.2-pro', name: 'GPT-5.2 Pro', desc: 'Version of GPT-5.2 that produces smarter and more precise responses.', logo: '/public/openai-logo.png', provider: 'openai', image: true, tools: true },
  { id: 'gpt-5-mini', name: 'GPT-5 Mini', desc: 'Faster, cost-efficient version of GPT-5 for well-defined tasks.', logo: '/public/openai-logo.png', provider: 'openai', image: true, tools: true },
  { id: 'gpt-5-nano', name: 'GPT-5 Nano', desc: 'Fastest, most cost-efficient version of GPT-5.', logo: '/public/openai-logo.png', provider: 'openai', tools: true },
  
  // OpenAI - GPT-4o Series
  { id: 'gpt-4o', name: 'GPT-4o', desc: 'High-intelligence model with vision, fast and affordable.', logo: '/public/openai-logo.png', provider: 'openai', image: true, tools: true },
  { id: 'gpt-4o-mini', name: 'GPT-4o Mini', desc: 'Small, fast, and affordable. Great for most tasks.', logo: '/public/openai-logo.png', provider: 'openai', image: true, tools: true },
  
  // OpenAI - o4 Reasoning Models (Latest)
  { id: 'o4', name: 'o4', desc: 'Latest reasoning model with enhanced capabilities.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true, image: true, tools: true },
  { id: 'o4-mini', name: 'o4-mini', desc: 'Faster reasoning model optimized for coding and math.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true },
  
  // OpenAI - o3 Reasoning Models
  { id: 'o3', name: 'o3', desc: 'Advanced reasoning model for complex multi-step tasks.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true, image: true, tools: true },
  { id: 'o3-mini', name: 'o3-mini', desc: 'Fast, flexible reasoning for coding and problem-solving.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true },
  
  // OpenAI - o1 Reasoning Models
  { id: 'o1', name: 'o1', desc: 'Original reasoning model for complex analysis.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true, image: true },
  { id: 'o1-mini', name: 'o1-mini', desc: 'Efficient reasoning model for coding and math.', logo: '/public/openai-logo.png', provider: 'openai', thinking: true },
  
  // Anthropic - Claude 4 Series (Latest)
    apiKeys: {
      groq: '', // Groq API key
      openai: '', // OpenAI API key
      anthropic: '', // Anthropic API key
      custom: '', // Custom provider API key (optional)

  groq: {
    name: 'Groq',
    baseUrl: 'https://api.groq.com/openai/v1',
    logo: '/public/groq-logo-orange.png',
    docsUrl: 'https://console.groq.com/docs',
    apiKeyPrefix: 'gsk_'
  },
  openai: {
    name: 'OpenAI',
    baseUrl: 'https://api.openai.com/v1',
    logo: '/public/openai-logo.png',
    docsUrl: 'https://platform.openai.com/docs',
    apiKeyUrl: 'https://platform.openai.com/api-keys',
    apiKeyPrefix: 'sk-'
  },
  { id: 'meta-llama/llama-4-maverick-17b-128e-instruct', name: 'Llama 4 Maverick 17B', provider: 'groq', desc: 'Powerful image understanding and analysis (600 t/s).', capabilities: { image: true, streaming: true }, preview: true },
  { id: 'meta-llama/llama-4-scout-17b-16e-instruct', name: 'Llama 4 Scout 17B', provider: 'groq', desc: 'Fast responses with image support (750 t/s).', capabilities: { image: true, streaming: true }, preview: true },
  { id: 'openai/gpt-oss-120b', name: 'GPT OSS 120B', provider: 'groq', desc: 'Flagship 120B parameter model with browser search and reasoning.', capabilities: { web: true, thinking: true, tools: true, streaming: true } },
  { id: 'openai/gpt-oss-20b', name: 'GPT OSS 20B', provider: 'groq', desc: 'Fast 20B model with web search, thinking, and code execution.', capabilities: { web: true, thinking: true, tools: true, streaming: true } },
  { id: 'moonshotai/kimi-k2-instruct-0905', name: 'Kimi K2', provider: 'groq', desc: 'Highly capable for writing and coding with 262K context.', capabilities: { streaming: true }, preview: true },
  { id: 'qwen/qwen3-32b', name: 'Qwen3-32B', provider: 'groq', desc: 'Multilingual responses with thinking capabilities.', capabilities: { thinking: true, streaming: true }, preview: true },
  
  // OpenAI models - GPT-5.2 Series (Latest)
  { id: 'gpt-5.2', name: 'GPT-5.2', provider: 'openai', desc: 'Most advanced model for coding and agentic tasks across industries.', capabilities: { image: true, streaming: true, tools: true } },
  { id: 'gpt-5.2-pro', name: 'GPT-5.2 Pro', provider: 'openai', desc: 'Version of GPT-5.2 that produces smarter and more precise responses.', capabilities: { image: true, streaming: true, tools: true } },
  { id: 'gpt-5-mini', name: 'GPT-5 Mini', provider: 'openai', desc: 'Faster, cost-efficient version of GPT-5 for well-defined tasks.', capabilities: { image: true, streaming: true, tools: true } },
  { id: 'gpt-5-nano', name: 'GPT-5 Nano', provider: 'openai', desc: 'Fastest, most cost-efficient version of GPT-5.', capabilities: { streaming: true, tools: true } },
  
  // OpenAI - GPT-4o Series
  { id: 'gpt-4o', name: 'GPT-4o', provider: 'openai', desc: 'High-intelligence model with vision, fast and affordable.', capabilities: { image: true, streaming: true, tools: true } },
  { id: 'gpt-4o-mini', name: 'GPT-4o Mini', provider: 'openai', desc: 'Small, fast, and affordable. Great for most tasks.', capabilities: { image: true, streaming: true, tools: true } },
  { id: 'chatgpt-4o-latest', name: 'ChatGPT-4o Latest', provider: 'openai', desc: 'Dynamic model continuously updated to current ChatGPT version.', capabilities: { image: true, streaming: true, tools: true } },
  
  // OpenAI - o4 Reasoning Models (Latest)
  { id: 'o4', name: 'o4', provider: 'openai', desc: 'Latest reasoning model with enhanced capabilities and configurable effort.', capabilities: { thinking: true, streaming: true, image: true, tools: true } },
  { id: 'o4-mini', name: 'o4-mini', provider: 'openai', desc: 'Faster reasoning model optimized for coding, math, and science.', capabilities: { thinking: true, streaming: true } },
  
  // OpenAI - o3 Reasoning Models
  { id: 'o3', name: 'o3', provider: 'openai', desc: 'Advanced reasoning model for complex multi-step tasks.', capabilities: { thinking: true, streaming: true, image: true, tools: true } },
  { id: 'o3-mini', name: 'o3-mini', provider: 'openai', desc: 'Fast, flexible reasoning for coding and problem-solving.', capabilities: { thinking: true, streaming: true } },
  
  // OpenAI - o1 Reasoning Models
  { id: 'o1', name: 'o1', provider: 'openai', desc: 'Original reasoning model for complex analysis.', capabilities: { thinking: true, streaming: true, image: true } },
  { id: 'o1-mini', name: 'o1-mini', provider: 'openai', desc: 'Efficient reasoning model for coding and math.', capabilities: { thinking: true, streaming: true } },
  { id: 'o1-pro', name: 'o1-pro', provider: 'openai', desc: 'Enhanced compute for hard problems.', capabilities: { thinking: true, streaming: true, image: true } },
  
  // Anthropic - Claude 4 Series (Latest)
    // Provider settings state
    providerSettings: {
      providers: {},   // { groq: { baseUrl, apiKey }, openai: {...} }
      models: [],      // User's configured models
      loaded: false
      var providerLogos = {
        groq: '/public/groq-logo-orange.png',
        openai: '/public/openai-logo.png',
        anthropic: 'vscode-icons:file-type-claude',
        custom: 'lucide:message-square'
      var logos = {
        groq: '/public/groq-logo-orange.png',
        openai: '/public/openai-logo.png',
        anthropic: 'vscode-icons:file-type-claude',
        custom: '/public/custom-logo.png'

                        />
                        <p class="text-xs text-muted-foreground mt-1">
                          The base OpenAI-compatible endpoint to use. See the <a href="https://console.groq.com/docs" target="_blank" class="text-blue-600 hover:underline">Groq documentation</a>.
                        </p>
                      </div>
                </div>
                
                <!-- OpenAI Provider Section -->
                <div class="surface" x-data="{ providerOpen: false }">
                  <button 
                    <span class="font-medium flex items-center gap-2">
                      <span class="iconify text-lg" data-icon="heroicons:chevron-right-20-solid" :class="{ 'rotate-90': providerOpen }" style="transition: transform 0.2s"></span>
                      <img src="/public/openai-logo.png" alt="OpenAI" class="w-5 h-5 object-contain" />
                      OpenAI
                    </span>
                  </button>
                            class="form-input flex-1"
                            placeholder="sk-..."
                            x-model="apiKeys.openai"
                            @change="updateApiKey('openai', $event.target.value)"
                            style="font-family: var(--font-mono, 'Menlo', monospace); font-size: 0.8125rem;"
                          />
                        <div class="flex items-center justify-between mt-1">
                          <p class="text-xs text-muted-foreground">
                            Get your API key from <a href="https://platform.openai.com/api-keys" target="_blank" class="text-blue-600 hover:underline">platform.openai.com</a>
                          </p>
                        </div>
                          type="text"
                          class="form-input"
                          x-model="providerSettings.providers.openai.baseUrl"
                          @input="updateProviderBaseUrl('openai', $event.target.value)"
                          style="font-family: var(--font-mono, 'Menlo', monospace); font-size: 0.8125rem;"
                        />
                        <p class="text-xs text-muted-foreground mt-1">
                          Default: https://api.openai.com/v1
                        </p>
                    </div>
                          <h4 class="text-sm font-semibold text-foreground mb-0">Models</h4>
                          <div class="flex items-center gap-1 text-xs">
                            <button type="button" class="btn-ghost btn-sm !px-2 !py-0.5 text-xs" @click="enableAllModels('openai')">All</button>
                            <span class="text-gray-300">/</span>
                            <button type="button" class="btn-ghost btn-sm !px-2 !py-0.5 text-xs" @click="disableAllModels('openai')">None</button>
                          </div>
                        </div>
                          type="button"
                          class="btn-secondary btn-sm"
                          @click="startAddingModel('openai')"
                          x-show="providerSettingsUI.addingModelToProvider !== 'openai'"
                        >
                          <span class="iconify mr-1" data-icon="lucide:plus"></span>
                      <!-- Add New Model Row -->
                      <div 
                        x-show="providerSettingsUI.addingModelToProvider === 'openai'" 
                        class="add-model-row flex items-center gap-2 p-3 mb-2 rounded-lg border"
                      >
                          placeholder="Enter model ID (e.g., gpt-4o)"
                          x-model="providerSettingsUI.newModelId"
                          :data-new-model-input="'openai'"
                          @keydown.enter.prevent="commitAddingModel()"
                          @keydown.escape.prevent="cancelAddingModel()"
                        
                        <div class="models-table-body expandable-table-body">
                          <template x-for="model in getModelsForProvider('openai')" :key="model.id">
                            <details class="expandable-row">
                              <summary class="text-sm">
                          </template>
                          
                          <div x-show="getModelsForProvider('openai').length === 0" class="p-4 text-center text-muted-foreground text-sm">
                            No models configured. Click "Add Model" to get started.
                          </div>
                    
                      <p class="text-sm text-muted-foreground">
                      Connect to any OpenAI-compatible API endpoint (e.g., Ollama, LM Studio, vLLM, Together AI).
                    </p>
                    
                        />
                        <p class="text-xs text-muted-foreground mt-1">
                          The base URL of your OpenAI-compatible API endpoint (must end with /v1)
                        </p>
                </div>

Use any unique string (e.g., "my-dnd-campaign", "solo-ironsworn-2024").
Supports: HTTP POST (OpenAI) and SSE (Claude.ai connectors)
Endpoint: ${url.origin}
`,
  }
  // Direct POST for HTTP transport (OpenAI style)
  if (req.method === "POST") {
    try {

Note: When changing a SQLite table's schema, change the table's name (e.g., add _2 or _3) to create a fresh table.
### OpenAI
```ts
import { OpenAI } from "https://esm.town/v/std/openai";
const openai = new OpenAI();
const completion = await openai.chat.completions.create({
  messages: [
    { role: "user", content: "Say hello in a creative way" },

import { generateObject } from "npm:ai";
import { openai } from "npm:@ai-sdk/openai";
import { z } from "npm:zod";
export async function summarizeContent(markdown: string, originalUrl: string): Promise<Summary> {
  const { object } = await generateObject({
    model: openai("gpt-4o"),
    schema: summarySchema,
    prompt: `You are summarizing web content that was saved as a task in Todoist.

  console.log("0.5 webhook");
  // Verify the webhook.
  const OPENAI_SIGNING_SECRET = Deno.env.get("OPENAI_SIGNING_SECRET");
  console.log("1 webhook");
  if (!OPENAI_SIGNING_SECRET) {
    console.error("🔴 webhook secret not configured");
    return c.text("Internal error", 500);
  console.log("2 webhook");
  const webhook = new Webhook(OPENAI_SIGNING_SECRET);
  const bodyStr = await c.req.text();
  let callId: string | undefined;

_2 or _3) to create a fresh table.
### OpenAI
```ts
import { OpenAI } from "https://esm.town/v/std/openai";
const openai = new OpenAI();
const completion = await openai.chat.completions.create({
  messages: [
    { role: "user", content: "Say hello in a creative way" },

import { OpenAI } from "https://esm.town/v/std/openai";
const enc = new TextEncoder();
  const payload = await req.json();
  const openai = new OpenAI({
    // Strongly recommended: store as a Val Town secret, not in code.
    // apiKey: Deno.env.get("BASETEN_API_KEY")!,
  };
  const openAIResponse = await openai.chat.completions.create(requestArgs);
  const stream = new ReadableStream<Uint8Array>({
      try {
        for await (
          const chunk of openAIResponse as unknown as AsyncIterable<any>
        ) {
          // Stop if the client disconnected