> ## Knowledge Base Index > Fetch the complete knowledge base index at: https://help.crisp.chat/sitemap.xml > Use this file to discover available pages before exploring further. > Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash). # How to create an AI Chatbot for Customer Service with Claude *Learn how to set up Hugo, the Crisp AI Agent, as a Claude-powered AI chatbot for customer support.* Hugo combines a supported **Anthropic Claude model** with your support content, business instructions, routing rules, and human handoff. This is sometimes described as a [Claude customer support chatbot,](https://crisp.chat/en/chatbot/claude/) but Hugo manages the support experience while Claude powers language understanding and answer generation. [Discover Hugo](https://hugo.ai/en/) to see how the AI Agent works. ||| Claude model availability evolves over time. Always check **AI Agent → Agent → Settings** for the current options in your workspace. --- # ${color}[#0080dd](How to set up a Claude AI chatbot for customer support) __To select an Anthropic Claude model in Hugo:__ 1. Sign in to [**Crisp**](https://app.crisp.chat). 2. Open **AI Agent → Agent → Settings**. 3. Under **Model**, disable **Choose automatically** and select a supported Claude model. 4. Review **Region**, **Enable Hugo's built-in fallback model**, **Allow agent to process images**, and **Allow agent to transcribe audio messages** when those options are available. When **Choose automatically** is enabled, Hugo may use a model from another provider. Select Claude manually if using Anthropic is an important part of your support configuration. ![Select an Anthropic Claude model in Hugo settings](https://storage.crisp.chat/users/helpdesk/website/-/8/7/a/e/87ae2703583ac800/cleanshot-2026-06-23-at-235234_13kl2n8.png =1000xauto) Choosing a model is only one part of the setup. Add reliable support resources from **AI Agent → Train**, define Hugo's behavior from **AI Agent → Guidance → Instructions**, and configure escalation or routing rules from **AI Agent → Guidance → Routing**. --- # ${color}[#0080dd](Why Claude works well for customer support) In our testing, Claude is one of the strongest model families for demanding customer support setups. It is particularly good at nuanced language, careful reasoning, and longer conversations, although its most capable models can consume more credits. __Claude models are especially useful for:__ * **Complex reasoning** → work through ambiguous requests, multi-step troubleshooting, and cases with several conditions * **Instruction following** → apply detailed business rules, support policies, and behavioral instructions consistently * **Source synthesis** → assemble information from interconnected articles, documents, and product resources * **Natural conversations** → maintain context and produce thoughtful, human-sounding replies across longer exchanges Claude can provide substantial reasoning horsepower, but the quality of your training resources still matters. Even the most capable model cannot fully compensate for outdated documentation, contradictory policies, or missing product information. --- # ${color}[#0080dd](Compare the Claude models available in Hugo) Hugo currently supports **Claude 4.8 Opus**, **Claude 4.6 Sonnet**, and **Claude 4.5 Haiku**, alongside selected legacy Claude models. The price levels below represent relative Hugo credit usage rather than a fixed price per conversation. | Claude model | Intelligence | Relative price | Recommended use | | ---- | | **Claude 4.8 Opus** | Very high | $$$ | Highly complex support environments with intricate instructions, interconnected sources, advanced routing rules, or requests requiring deep reasoning | | **Claude 4.6 Sonnet** | High | $$ | Most advanced customer support setups, offering an excellent balance between answer quality, reasoning, and credit usage | | **Claude 4.5 Haiku** | Medium | $ | Clear documentation, straightforward support questions, higher conversation volumes, and cost-conscious deployments | **Claude Opus** offers the most horsepower, but most support teams do not need that much power for every conversation. **Claude Sonnet** is usually the best cost-to-performance balance and is also the model Crisp uses for its own customer support setup. **Claude Haiku** is a practical starting point when your documentation and support processes are already clear. | See [which AI models Hugo supports and how to configure them](https://help.crisp.chat/en/article/how-to-choose-and-configure-hugos-ai-model-1m4mm57/) for the latest model, region, and relative pricing options. --- # ${color}[#0080dd](Test and launch your Claude customer support chatbot) Before going live, open **AI Agent → Evaluate → Playground → Compare model options**. Test each Claude model with the same real customer questions, including simple FAQs, ambiguous requests, longer conversations, multi-source questions, and situations that should trigger a routing rule or human escalation. Compare **answer accuracy**, **instruction following**, **source usage**, **reasoning quality**, **response speed**, and **credit usage**. Opus may perform best on your hardest questions while Sonnet or Haiku provides better overall value across everyday conversations. Once you have selected the right configuration, deploy Hugo from **AI Agent → Agent → Activation**. For the complete training, guidance, testing, and activation process, follow [Getting started with Hugo AI Agent](https://help.crisp.chat/en/article/getting-started-with-hugo-ai-agent-w6gbux/). --- # ${color}[#0080dd](Frequently Asked Questions) *Still have questions which were not covered in this article? Here is a collection of the most frequently asked questions on this topic.* ###### ${color}[#F08820](Do I need an Anthropic API key to use Claude with Hugo?) No. Supported Claude models are managed directly through Hugo, and you cannot connect an external Anthropic API key. Select the model you want from **AI Agent → Agent → Settings**. ###### ${color}[#F08820](Does Hugo support every Claude model released by Anthropic?) Almost. Hugo exposes a selected list of supported models rather than every Claude release. The models displayed under **AI Agent → Agent → Settings** are the source of truth for your workspace. At Crisp, we regularly evaluate and implement new model releases so Hugo continues to offer modern and competitive options. ###### ${color}[#F08820](Is Claude Opus always the best choice?) Not necessarily. **Claude Opus** offers the most reasoning horsepower, but it also consumes more credits and may be excessive for everyday support. **Claude Sonnet** provides a strong balance for most advanced setups, while **Claude Haiku** is often sufficient for clear documentation and straightforward customer questions.