Choosing the Right AI Model
ChatbotIQ supports models from four providers: OpenAI, Anthropic, Google, and Mistral. Each model offers a different balance of quality, speed, and cost. This guide helps you pick the right one.
The models at a glance
Section titled “The models at a glance”| Model | Provider | Quality | Speed | Best for |
|---|---|---|---|---|
| GPT-5 Mini | OpenAI | Good | Fast | General use. The best default choice. |
| Gemini 3 Flash | Good | Very fast | Speed-critical applications. | |
| Mistral Small 3.2 | Mistral | Good | Fast | Lightweight tasks. EU-hosted. |
| Mistral Medium 3.5 | Mistral | Very good | Moderate | Multilingual reasoning, EU-hosted. |
| GPT-5.2 | OpenAI | Excellent | Moderate | Complex technical questions. |
| Gemini 3 Pro | Very good | Moderate | Multilingual and complex tasks. | |
| Claude Sonnet 4.6 | Anthropic | Excellent | Moderate | Nuanced, well-structured responses. |
| Claude Opus 4.6 | Anthropic | Best | Slower | Highest quality for critical use cases. |
For credit costs per model, see Plans, Credits & Pricing.
How to think about model choice
Section titled “How to think about model choice”Start with GPT-5 Mini
Section titled “Start with GPT-5 Mini”For most use cases, GPT-5 Mini is the right choice. It’s fast, accurate, and costs 1x credits — meaning your monthly credit budget goes furthest. It handles documentation queries, support questions, and general knowledge retrieval very well.
Only upgrade if you have a specific reason.
When to upgrade
Section titled “When to upgrade”Choose a more powerful model if:
- Your content is highly technical — complex engineering documentation, legal text, or scientific material benefits from models with stronger reasoning (GPT-5.2, Claude Sonnet 4.6).
- You need nuanced responses — if your bot handles sensitive topics (HR policies, medical information), a model like Claude Sonnet 4.6 produces more careful, well-structured answers.
- Quality is more important than cost — for customer-facing bots where every answer matters, investing in a higher-quality model can be worth it.
- Multilingual content — Gemini 3 Pro and Claude models tend to perform better across languages.
Stay with GPT-5 Mini if:
- Your content is straightforward (help articles, FAQs, product docs).
- You’re on a budget and want maximum messages per credit.
- Speed matters more than nuance.
Quality vs. cost tradeoff
Section titled “Quality vs. cost tradeoff”Here’s a practical way to think about it:
| Scenario | Recommended model | Why |
|---|---|---|
| Help center / FAQ bot | GPT-5 Mini | Simple Q&A, high volume, cost matters |
| Technical documentation bot | GPT-5 Mini or GPT-5.2 | Try Mini first; upgrade if answers lack depth |
| Customer support (external-facing) | GPT-5 Mini or Claude Sonnet 4.6 | Mini for volume; Claude for quality |
| Internal knowledge bot | GPT-5 Mini | Team members tolerate less polish |
| Legal / compliance bot | Claude Sonnet 4.6 | Nuance and accuracy are critical |
| Small site, low volume | Claude Opus 4.6 | Few messages = cost doesn’t matter, maximize quality |
Related
Section titled “Related”- Plans, Credits & Pricing — model pricing
- Choose and Switch AI Models — step-by-step switching guide
- Bot Settings Reference — all model-related settings