Log in

Choose your bot's AI model

Pick the AI model that powers your bot's replies. Models differ in quality, speed, what they support and how many message tokens they use.

ℹ️

Plan requirement

Choosing a model is available on the Standard, Pro and Enterprise plans. On Free and Basic, bots use the default model. See plans.

Who can do this: team owners and admins.

Available models

Anthropic

Name
Context window
Power ups
Vision
Claude Haiku 3 200,000
Claude Haiku 4.5 200,000
Claude Sonnet 5 1,000,000
Claude Sonnet 4.6 1,000,000
Claude Sonnet 4.5 1,000,000
Claude Sonnet 4 1,000,000
Claude Opus 4.6 1,000,000
Claude Opus 4.7 1,000,000
Claude Opus 5 1,000,000
Claude Opus 4.5 200,000
Claude Opus 4.8 1,000,000
Claude Opus 4.8 (Fast) 1,000,000
Claude Opus 5 (Fast) 1,000,000
Claude Opus 4 200,000
Claude Opus 4.1 200,000
Claude Opus 4.7 (Fast) 1,000,000

Cohere

Name
Context window
Power ups
Vision
Cohere - Command R 128,000
Command A 256,000
Cohere - Command R+ 128,000

DeepSeek

Name
Context window
Power ups
Vision
DeepSeek V4 Flash 1,048,576
DeepSeek V3 163,840
DeepSeek V4 Pro 1,048,576
DeepSeek R1 64,000

Google

Name
Context window
Power ups
Vision
Google - Gemini 2.5 Flash Lite 1,048,576
Gemini 3.1 Flash Lite 1,048,576
Gemini 3.5 Flash Lite 1,048,576
Google - Gemini 2.5 Flash 1,048,576
Gemini 3.7 Flash 1,048,576
Google - Gemini 3 Flash 1,048,576
Google - Gemini 2.5 Pro 1,048,576
Gemini 3.6 Flash 1,048,576
Gemini 3.5 Flash 1,048,576
Google - Gemini 3.1 Pro 1,048,576

Meta

Name
Context window
Power ups
Vision
Llama 4 Scout 327,680
Llama 4 Maverick 1,048,576

Mistral

Name
Context window
Power ups
Vision
Mistral - Mistral Small 32,768
Mistral Large 3 262,144
Mistral Medium 3.5 262,144
Mistral - Mistral Large 128,000
Mistral - Open Mixtral 8x22b 65,536

MoonshotAI

Name
Context window
Power ups
Vision
Kimi K2 131,072
Kimi K2.6 262,144
Kimi K3 1,048,576

OpenAI

Name
Context window
Power ups
Vision
GPT-5 Nano 400,000
GPT-5.6 Luna ourdefault1,050,000
GPT-5.6 Luna Pro 1,050,000
GPT-4.1 Nano 1,047,576
GPT-4o Mini 128,000
GPT-5.4 Nano 400,000
GPT-5 Mini 400,000
GPT-4.1 Mini 1,047,576
GPT-3.5 Turbo 16,385
GPT-5.4 Mini 400,000
GPT-5.6 Terra 1,050,000
GPT-5.6 Terra Pro 1,050,000
GPT-5.1 400,000
GPT-5.1 Chat 128,000
GPT-5 400,000
GPT-5.2 400,000
GPT-4.1 1,047,576
GPT-5.4 1,050,000
GPT-4o 128,000
GPT-5.6 Sol 1,050,000
GPT-5.6 Sol Pro 1,050,000
GPT-5.5 1,050,000
GPT-4 Turbo 128k 128,000
GPT-5 Pro 400,000
GPT-5.2 Pro 400,000
GPT-5.4 Pro 1,050,000
GPT-5.5 Pro 1,050,000
GPT-4 8,191

Perplexity

Name
Context window
Power ups
Vision
Sonar 127,072
Sonar Pro 200,000

Qwen

Name
Context window
Power ups
Vision
Qwen3.7 Flash 1,000,000

xAI

Name
Context window
Power ups
Vision
Grok 4.3 1,000,000
Grok 4.20 2,000,000
Grok 4.6 500,000
Grok 4.5 500,000

Z.ai

Name
Context window
Power ups
Vision
GLM 5.3 Flash 1,310,720
GLM 4.7 202,752
GLM 4.6 202,752
GLM 4.5 131,000
GLM 5.1 202,752
GLM 5.2 1,048,576
  • Context window: how much text (prompt, knowledge and conversation history) the model can take in at once.
  • Power ups: whether the model can use power-ups. If a model doesn't support them, any power-ups on the bot won't work.
  • Vision: whether the model can read images. Only these models can use Enable file upload (multi-modal).

How to choose

  • Start with the default model and only change it if testing shows you need to.
  • Need power-ups (booking, handoff, calling your API)? Pick a model that supports them.
  • Want users to send images? Pick a model with vision.
  • Complex questions, long instructions or multi-step power-up workflows? Try a more capable model and compare the answers with test cases.
  • High volume, simple questions? A smaller model uses fewer message tokens per reply.

Model cost

More expensive models use more message tokens per reply. Each model has an input and output multiplier; when you pick a model, a note below the selector shows them ("This model uses …x input tokens and …x output tokens"). See message tokens for how multipliers work and the full pricing table.

Change the model

  1. Open your bot and go to the General settings tab.
  2. Scroll to Model settings (advanced).
  3. Open the Model selector and choose a model.
  4. Check the notes below the selector: the token multiplier, and a warning if the model doesn't support power-ups.
    The Model selector in Model settings (advanced), with a note below it showing the model's input and output token multipliers
  5. Click Update bot.

Changing the model can change the maximum for Max context size (base tokens). If your setting was at the old model's maximum, it moves to the new model's maximum.

Check it works

Ask your bot a few typical questions and compare the answers and speed with the previous model. Test cases let you run the same questions against each model and compare the results.

Troubleshooting

The model selector is greyed out

Your plan doesn't include model selection. The message "Model selection is only available on Standard & Pro plans" appears below it. Upgrade to Standard or above to choose a model.

My power-ups stopped working after changing model

The new model doesn't support power-ups. Choose a model with a tick in the Power ups column.

Related

Last updated