Deep Dive

Building the Best Free AI Chatbot: 16 Modes, 18 APIs, Zero Cost

Most AI chatbots do one thing — answer questions. This one does sixteen things, pulls data from eighteen free APIs, and costs nothing to use. Here’s how every mode works and why it exists.

Try It Free
16
Intelligent Modes
18
Free APIs
$0
Cost to Use
0
Accounts Required

The goal: a chatbot that actually does things

Most free chatbots are wrappers around a single LLM call. You type something, the model generates a response, and that’s it. No real data. No live information. No ability to generate images, solve math symbolically, or build you a working game.

The goal with helloandy.net’s AI Chat was different: build a chatbot that genuinely does useful things across many domains — weather, code, math, research, image generation, currency conversion, trivia — and make it completely free. No account, no API key, no credit card.

The result is a 16-mode chatbot powered by 18 free APIs with a smart LLM router that automatically detects what you’re asking and routes your question to the right mode. You don’t pick modes manually. Just ask naturally and the system figures it out.

How the smart router works

At the core is an LLM-powered query classifier. When you send a message, the router reads your input and classifies it into one of 16 categories with a single fast LLM call. The router uses zero-temperature inference for consistency — the same question should always route the same way.

The router prompt contains detailed classification rules with edge cases. For example, “how many feet in a mile?” routes to calculate (not lookup), while “what does HN think about X?” routes to research (not news). These distinctions matter because each mode has different API calls, prompt engineering, and output formats.

After classification, the system dispatches to the appropriate handler. Some modes are direct LLM calls. Others involve multi-step pipelines with external API calls, data enrichment, and synthesis. Here’s every mode in detail.


All 16 modes explained

💬
Chat
Casual conversation, greetings, small talk, and questions about the assistant. Warm, friendly responses under 60 words. Invites users to explore other modes when they say hello.
API: OpenRouter (direct LLM)
🌤️
Weather
Real-time weather for any location on Earth. Shows current temperature (°C and °F), feels-like, humidity, wind speed and direction, visibility, plus a 3-day forecast with UV index.
API: wttr.in (weather data)
🔢
Calculate
Everyday arithmetic, percentages, compound interest, date math, and unit conversions. Shows structured steps: Expression → Steps → Answer with units. Zero temperature for precision.
API: OpenRouter (direct LLM, T=0.0)
📐
Math
Formal academic mathematics — algebra, calculus, linear algebra, combinatorics, statistics, geometry, proofs. Runs SymPy symbolic computation first, then expands with step-by-step LLM explanation.
API: SymPy (symbolic computation) + OpenRouter
💻
Code
Programming help, debugging, algorithm design, and language-specific syntax questions. Covers Python, JavaScript, SQL, Git, Bash, CSS, and more. Includes a format-checker that validates output structure.
API: OpenRouter (direct LLM, T=0.2)
📰
News
Current events and recent developments. Searches DuckDuckGo and Hacker News for time-sensitive information, then synthesizes findings into a coherent summary with source links.
API: DuckDuckGo + Hacker News (Algolia) + OpenRouter
🔍
Lookup
Quick factual answers with a single definitive answer. “Who invented X?” “Capital of Y?” Uses Wikipedia and DuckDuckGo for fast, authoritative responses.
API: Wikipedia + DuckDuckGo + OpenRouter
🔬
Research
The flagship mode. Deep multi-source synthesis for complex questions. Expands queries into 3 sub-queries, searches 10+ sources (academic papers, Wikipedia, web, GitHub, Hacker News), then synthesizes with mandatory citations.
API: DuckDuckGo, Wikipedia, arXiv, Semantic Scholar, Crossref, GitHub, Hacker News, Open Library, Wikidata, World Bank + OpenRouter
🎨
Image
AI image generation from text prompts. Uses the FLUX model via Pollinations.ai to create illustrations, art, photos, and visual concepts. No account needed — completely free generation.
API: Pollinations.ai (FLUX model)
📱
QR Code
Generate QR codes for URLs, text, WiFi credentials, or any data. Uses LLM to extract the right payload from natural language requests, then generates a scannable QR code image.
API: QR Server (qrserver.com) + OpenRouter
🌐
HTML
Generate beautiful, complete HTML pages on demand. Landing pages, forms, dashboards, error pages, newsletters, pricing tables — with full CSS, responsive design, and lint validation with auto-retry.
API: OpenRouter (T=0.25, 3500 tokens) + HTML linter
🎮
Game
Create playable browser games from a description. Snake, Tetris, Pong, memory matching, card games — generates complete HTML/CSS/JavaScript with canvas or DOM rendering. Validated with lint checks and auto-retry.
API: OpenRouter (T=0.3, 4000 tokens) + HTML/JS linter
📊
Data
Live economic and government statistics from specialized databases. Current unemployment rate, inflation, GDP, interest rates — pulls real numbers from the Federal Reserve Economic Data (FRED) API.
API: FRED (Federal Reserve) + OpenRouter
💱
Currency
Real-time currency conversion using European Central Bank exchange rates. Supports all major and many minor currencies. Shows the rate, converted amount, and last-updated timestamp.
API: Frankfurter (ECB data)
📖
Word
Dictionary definitions, pronunciation, etymology, parts of speech, and usage examples. Pulls from the Free Dictionary API for comprehensive word information.
API: Free Dictionary API (dictionaryapi.dev)
🧠
Trivia
Random trivia questions across categories — science, history, geography, entertainment, sports, and more. Multiple choice or true/false. Pulls from the Open Trivia Database.
API: Open Trivia DB (opentdb.com)

The research engine: where depth matters

Research mode is the most complex and capable mode. It’s what sets this chatbot apart from simple LLM wrappers. Here’s the full pipeline:

  1. Query expansion — Your question gets expanded into 3 semantically diverse sub-queries using Jaccard similarity filtering. This ensures broader source coverage than a single search.
  2. Source detection — The system analyzes your query to determine which APIs to search. Academic questions hit arXiv and Semantic Scholar. GitHub questions search repositories. Community-opinion questions go to Hacker News.
  3. Multi-source search — Searches run in parallel across up to 10 sources: DuckDuckGo, Wikipedia, arXiv, Semantic Scholar, Crossref, GitHub, Hacker News, Open Library, Wikidata, and World Bank.
  4. Credibility scoring — Every source gets a credibility score. Academic papers (arXiv, Semantic Scholar) score highest. Government and .edu domains rank above commercial sites. Wikipedia is high but not top.
  5. Forced-attribution synthesis — The top sources are extracted into structured fact blocks. The LLM must cite every single fact — there is no way to write an uncited claim. This produces answers with 7+ real citations in every response.

The result: research answers that cite real papers, real data, and real community discussions. Not hallucinated references — verified, clickable sources from authoritative databases.

Quality engineering: the ratchet process

Building a chatbot is easy. Making it consistently good is hard. The quality of this chatbot was systematically improved through an iterative process called the ratchet — a series of measured improvements where each iteration must score higher than the last.

The evaluation framework tests 10 diverse questions across 5 runs each, producing 50 data points per iteration. Every answer is scored on:

Over 46 iterations, the median composite score improved from 5.2 to 8.12 — a 56% improvement through systematic source injection, citation engineering, prompt architecture, and synthesis pipeline redesign.

Key insight: The biggest quality gains came not from model upgrades or prompt tweaking, but from source injection — pre-loading authoritative facts into the synthesis pipeline so the model has real data to cite. Injecting 2 curated sources for a topic typically improves scores by 1.0–2.0 points.

The 18 free APIs

Every API used by this chatbot is completely free. No API keys required for most — only OpenRouter (for LLM inference) and FRED (for economic data) need keys, and both offer free tiers.

# Service Purpose Used By
1OpenRouterLLM inference (all modes)All modes
2wttr.inReal-time weather dataWeather
3Pollinations.aiAI image generation (FLUX)Image
4QR ServerQR code generationQR
5FREDFederal Reserve economic dataData
6FrankfurterECB currency exchange ratesCurrency
7Free DictionaryWord definitionsWord
8Open Trivia DBTrivia questionsTrivia
9DuckDuckGoWeb searchResearch, News, Lookup
10WikipediaEncyclopediaResearch, Lookup
11Hacker NewsTech communityResearch, News
12arXivAcademic papersResearch
13Semantic ScholarAcademic searchResearch
14CrossrefDOI/citation lookupResearch
15GitHubRepository searchResearch
16Open LibraryBook searchResearch
17WikidataStructured knowledgeResearch
18World BankDevelopment dataResearch

Bring your own key (BYOK)

The chatbot works out of the box with no configuration — it uses a shared free-tier OpenRouter API key. But if you want faster responses and no rate limiting, you can bring your own OpenRouter key.

Click the settings icon in the chat interface, paste your API key, and it’s stored locally in your browser. Your key is sent directly to OpenRouter — it’s never stored on the server. The free model (arcee-ai/trinity-large-preview) still costs nothing even with your own key, but you bypass the shared 15-requests-per-hour limit.

Format validation and self-correction

Several modes include a format-checker pipeline that validates LLM output structure and retries if needed. The calculate mode checks for the required Expression → Steps → Answer format. The HTML and game modes run a full lint pass — checking for matched tags, required elements, valid JavaScript, and responsive design patterns. If validation fails, the system provides specific error feedback and regenerates.

This means the chatbot can catch and fix its own mistakes before you see them. A game that crashes on load gets rebuilt. A calculation missing its answer line gets reformatted. Self-correction produces noticeably better output quality than raw single-pass generation.

What makes this different

There are thousands of AI chatbot wrappers. Here’s what this one does differently:

Try it now: Visit helloandy.net/ai-chat to try all 16 modes. Ask about the weather, generate an image, solve a math problem, or ask a deep research question. No account needed. Or check out the Skill Generator and other free developer tools.

helloandy.net provides free AI tools and tutorials for developers. No account required.

Related Guides