> For the complete documentation index, see [llms.txt](https://docs.wonderchat.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.wonderchat.io/setup-guides/setting-up-your-chatbot/selecting-between-ai-models-for-your-chatbots.md). # Selecting Between AI models for your chatbots A guide to decide how you can select the best AI model out of the 10 different AI model selections to power your AI chatbot. We support a number of different models provided by multiple AI service providers, namely OpenAI, Anthropic, Google, DeepSeek, Moonshot AI, and Microsoft Azure, each with their own series of models. Each model has its own unique specifications and use cases, and in this guide we will explain how to choose the best model for your use case. *** ## Selecting Between OpenAI models for your Chatbots Choosing the right OpenAI model is crucial for optimizing your chatbot's performance. OpenAI offers models with varying capabilities, from fast and affordable everyday chat to high-reasoning models for complex, multi-step tasks. This guide will help you understand the key differences in performance, token limits, and use cases, so you can select the best model for your needs — whether it's for simple queries or complex tasks. ### **1. GPT-5.6 Luna** | Attribute | Details | | ----------- | --------------------------------------------------------------------------------------------------------------------------- | | Description | The new default OpenAI model — fast, affordable, and built for everyday chatbot conversations. | | Costs | 1 message per user query and chat response | | Strengths | Best all-round balance of speed, cost, and quality. The recommended starting point for most chatbots. | | Use Cases | General-purpose support bots, FAQ bots, and any chatbot where cost efficiency matters without sacrificing response quality. | ### **2. GPT-5.6 Terra** | Attribute | Details | | ----------- | ----------------------------------------------------------------------------------------------------------- | | Description | A balanced GPT-5.6 model - GPT-5.5-class quality at roughly half the flagship cost. | | Costs | 5 messages per user query and chat response | | Strengths | Stronger reasoning than Luna while still being meaningfully cheaper than the flagship Sol model. | | Use Cases | Chatbots that need better handling of nuanced or multi-part questions but don't require top-tier reasoning. | ### **3. GPT-5.6 Sol** | Attribute | Details | | ----------- | ------------------------------------------------------------------------------------------------------------------------------------ | | Description | The flagship of the GPT-5.6 family — top-tier reasoning for complex and agentic tasks. | | Costs | 10 messages per user query and chat response | | Strengths | Highest intelligence in the OpenAI lineup we offer. Best choice when accuracy on complex or multi-step tasks matters more than cost. | | Weaknesses | Most expensive OpenAI option available. | | Use Cases | Complex research assistants, advanced troubleshooting bots, and any chatbot performing multi-step reasoning or agentic tool use. | ### **4. GPT-5.5** | Attribute | Details | | ----------- | ------------------------------------------------------------------------------------------------------- | | Description | A high-reasoning OpenAI model for complex, multi-step tasks. | | Costs | 10 messages per user query and chat response | | Strengths | Strong reasoning, comparable cost tier to GPT-5.6 Sol. | | Use Cases | Similar use cases to GPT-5.6 Sol — consider Sol first unless you have a specific reason to stay on 5.5. | ### **5. GPT-4o** | Attribute | Details | | ----------- | ----------------------------------------------------------------------------------------------- | | Description | A flexible legacy OpenAI model, strong at multilingual tasks. Requires at least the Basic plan. | | Costs | 5 messages per user query and chat response | | Strengths | Solid multilingual performance, well-established and stable. | | Weaknesses | Superseded in reasoning quality by the GPT-5.6 family at a similar cost tier. | | Use Cases | Multilingual support chatbots, or existing setups already tuned around GPT-4o behavior. | #### In Summary: * If you're getting started or want the best cost-to-quality ratio, **GPT-5.6 Luna** (the current default) is the right choice for most chatbots. * If you need stronger reasoning without paying flagship prices, **GPT-5.6 Terra** is a good middle ground. * If your chatbot needs to reliably handle complex, multi-step, or agentic tasks, use **GPT-5.6 Sol**. * If you specifically need strong multilingual support and are comfortable on a legacy-but-supported model, **GPT-4o** remains available. \**GPT-4, GPT-4 Turbo, GPT-4o Mini, GPT-4.1, and GPT-4.1 Mini have been retired from selection for new setups. GPT-4 and GPT-4 Turbo are scheduled to be fully shut down on **2026-10-23** — if your chatbot is still on one of these, switch it to a GPT-5.6 model before then.* *** ## Selecting Between Claude AI Models Anthropic's current lineup available on Wonderchat spans Claude Haiku, Claude Sonnet, and Claude Opus — each targeting a different balance of speed, cost, and reasoning depth. ### **1. Claude Opus** | Attribute | Details | | ----------- | ----------------------------------------------------------------------------------------------------------------------------- | | Description | The most powerful Anthropic model available, with superior reasoning for complex tasks. Requires at least the Scale plan. | | Costs | 10 messages per user query and chat response | | Strengths | Exceptional performance on highly complex tasks with near-human fluency and understanding. | | Weaknesses | Most expensive Claude model in our lineup. | | Use Cases | Advanced research tools for complex documents, critical analysis of large report corpora, and high-stakes accuracy use cases. | ### **2. Claude Sonnet** | Attribute | Details | | ----------- | ------------------------------------------------------------------------------------------------------------------- | | Description | Anthropic's latest Sonnet model — near-Opus quality with balanced speed and cost. Requires at least the Scale plan. | | Costs | 5 messages per user query and chat response | | Strengths | Strong reasoning at a meaningfully lower cost than Opus; the recommended default Claude model. | | Use Cases | General-purpose chatbots that want Claude's tone and knowledge-base handling. | ### **3. Claude Haiku** | Attribute | Details | | ----------- | -------------------------------------------------------------------------------------- | | Description | Fast, affordable Anthropic model with strong capabilities for its price point. | | Costs | 2 messages per user query and chat response | | Strengths | Most cost-effective Claude model, near-instant responsiveness. | | Weaknesses | Less suited to deep, multi-step reasoning than Sonnet or Opus. | | Use Cases | FAQ bots on a simple knowledge base that need quick answers without in-depth analysis. | In summary, Claude remains a popular alternative to OpenAI's GPT models — some prefer it simply performs better against their specific knowledge base. * If you're working on a budget, **Claude Haiku** is the most cost-effective choice. * For most chatbots, **Claude Sonnet** is the recommended default — strong reasoning at a fraction of Opus' cost. * If you need maximum reasoning depth for complex research or analysis, use **Claude Opus**. > **Note:** Claude 3 Opus/Sonnet/Haiku, Claude 3.5/3.7/4/4.5 Sonnet, and Claude Opus 4.6 deprecated from selection for new setups — existing chatbots on these models keep working, but new configurations should use the models above. ## Selecting Between Gemini Models Google's current Gemini lineup on Wonderchat covers a fast/cheap tier and two higher-reasoning tiers. ### **1. Gemini 3.1 Flash Lite** | Attribute | Details | | ----------- | ------------------------------------------------------------------------------------------- | | Description | The cheapest, fastest Gemini model available. | | Costs | 1 message per user query and chat response | | Strengths | Best for high-volume, latency-sensitive use cases. | | Use Cases | Simple FAQ-style chatbots and high-traffic deployments where cost per message matters most. | ### **2. Gemini 3.1 Pro** | Attribute | Details | | ----------- | -------------------------------------------------------------------------------------------------------------------------------------- | | Description | Google's advanced Gemini Pro model with strong reasoning and a very large (1M-token) context window. Requires at least the Scale plan. | | Costs | 5 messages per user query and chat response | | Strengths | Long context window, strong reasoning for detailed or research-heavy tasks. | | Use Cases | Chatbots that need to reason over very large amounts of context, or detailed multi-part analysis. | ### **3. Gemini 3.5 Flash** | Attribute | Details | | ----------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | Description | Google's newest Gemini model — frontier-level intelligence that beats Gemini 3.1 Pro on coding and agentic tasks while running faster. Requires at least the Scale plan. | | Costs | 2 messages per user query and chat response | | Strengths | Best combination of speed and intelligence currently offered by Google. | | Use Cases | Agentic or coding-adjacent chatbot tasks, and any use case where you'd otherwise reach for 3.1 Pro but want faster responses. | *** ## Other AI Models: DeepSeek, Moonshot, and Azure ### 1. DeepSeek V4 Flash | Attribute | Details | | ----------- | --------------------------------------------------------------------------------------- | | Description | DeepSeek's newest fast, extremely affordable model with a large context window. | | Costs | 1 message per user query and chat response | | Strengths | Very low cost per message combined with a generous context window. | | Use Cases | High-volume, cost-sensitive chatbots that still need a reasonably large context window. | ### 2. DeepSeek V4 Pro | Attribute | Details | | ----------- | ----------------------------------------------------------------------------------------------------------------------------------- | | Description | DeepSeek's newest flagship model - most capable with strong reasoning and a large context window. Requires at least the Basic plan. | | Costs | 5 messages per user query and chat response | | Strengths | Strong reasoning at a lower cost than comparable OpenAI/Anthropic tiers. | | Use Cases | Chatbots that want stronger reasoning without moving up to Scale-tier-gated models. | ### 3. Azure GPT-4.1 Mini / GPT-4o Mini | Attribute | Details | | ----------- | --------------------------------------------------------------------------------------------------------------------------------------------------------- | | Description | OpenAI models hosted on Microsoft Azure infrastructure, for customers with Azure data-residency or compliance requirements. Requires the Enterprise plan. | | Costs | 1 message per user query and chat response | | Strengths | Fast, low-cost, with the compliance benefits of Azure hosting. | | Use Cases | Enterprise customers with Azure-specific compliance or data-residency requirements. | We will be continually updating this page as we introduce more and more AI models. If you would like to request for a custom AI model to be made available on Wonderchat, you can reach out to provide us [**feedback here**](https://wonderchat.featurebase.app/). *** If you have any more questions, feel free to reach out to us at ****