Model

Choose which LLM powers your agent — including PolyAI Raven, our proprietary model family built for conversational AI.

Choose the LLM that powers your agent. The model directly affects response quality, latency, and cost — a wrong choice can make your agent slow, expensive, or inaccurate.

Recommended: PolyAI’s Raven model family is purpose-built for conversational AI and is the recommended choice for most deployments. Raven supports 24+ languages, delivers sub-300ms latency, and produces concise, natural responses without extensive prompt engineering. Raven 3.5 powers both voice and chat, with auto-reasoning and out-of-domain detection.

Available models

PolyAI Raven (Recommended)

PolyAI’s proprietary Raven model family is specialized for customer service across voice and chat. Raven eliminates the trade-off between speed and accuracy — delivering sub-300ms latency, fewer errors, and more natural responses than general-purpose LLMs.

Raven 3.5 is the recommended model for all deployments. It supports voice and chat, 24+ languages, auto-reasoning, out-of-domain detection, custom style following, and built-in safety.

Learn more about Raven

Full details on capabilities, supported languages, and why Raven is recommended for most deployments.

OpenAI

Model	Best for
GPT-5.2	High-quality interactions requiring nuance and strong reasoning
GPT-5.2 chat	Extended dialogue and conversational stability
GPT-5 mini	Lower latency and reduced cost for mid-complexity use cases
GPT-5 nano	Simple tasks and fast-response workloads
GPT-4o	Versatile balance of reasoning, speed, and cost
GPT-4o mini	Everyday queries and high-volume deployments
GPT-4.1	Strong reasoning with improved cross-task performance
GPT-4.1 mini	Cost-effective, latency-focused for lighter workloads
GPT-4.1 nano	Minimal compute and high throughput

See OpenAI model documentation for detailed specifications.

Amazon Bedrock

Model	Best for
Claude 3.5 Haiku	Simple, predictable tasks with strong safety alignment
Nova Micro	Efficiency with strong general-purpose performance

See Anthropic Claude docs and Amazon Nova docs for more details.

Configuring the model

1. Open model settings

   Navigate to Channels > Voice > Voice configuration or Channels > Chat > Chat configuration to select the model for each channel.

2. Select a model

   Choose the desired model from the dropdown.

3. Save changes

   Click Save to apply your changes.

Choosing the right model

For most use cases, start with Raven — it is optimized for the PolyAI platform and delivers the best out-of-the-box experience for both voice and chat. General-purpose models like GPT and Claude require extensive prompting to work well for customer service conversations. With Raven, the right conversational behavior is built in.

Scenario	Recommended model
Voice agent	Raven 3.5
Chat agent	Raven 3.5
Voice + chat (multichannel)	Raven 3.5 for both channels
Multilingual agent	Raven 3.5 — 24+ languages with near-perfect language consistency
Cost-sensitive, simple tasks	GPT-5 nano or GPT-4.1 nano
Specific compliance requirements	Evaluate based on your provider requirements

Related pages

Raven

Full details on PolyAI’s proprietary model family — capabilities, versions, and supported languages.

Bring your own model

Connect your own LLM endpoint to PolyAI.

Voice configuration

Select the model for your voice channel.

Chat configuration

Select the model for your chat channel.