Latest Frontier Model Support and More in Releases 2026.21/22

Newest OpenAI, Anthropic, and Mistral AI Models

You can now build with state-of-the-art models from OpenAI, Anthropic, and Mistral AI in Cognigy, and choose the best one for each use case.

OpenAI and Microsoft Azure OpenAI

• gpt-6-sol and gpt-6.1-sol: complex tasks, such as coding.
• gpt-6-luna: lower-cost model for high-volume tasks with a clear goal, such as extracting information or answering quick questions.

Anthropic

• claude-opus-5-5: complex work that requires careful judgment.
• claude-sonnet-5-5: well-scoped everyday tasks that need fast output.
• claude-haiku-5-5: high-volume, cost-sensitive applications.

Mistral AI

• ministral-14b-latest: higher accuracy at a low cost.
• ministral-8b-latest: compact model with a strong performance-to-cost ratio.
• ministral-3b-latest: fast, lightweight tasks.

Using larger models only where the task demands it keeps quality high where it counts and lowers cost and latency everywhere else.

Amazon Polly TTS Engine Selection in Voice Gateway

Voice Gateway now lets you choose between the Generative, Neural, and Standard Amazon Polly engines, with a separate engine for the fallback speech vendor. 

That gives you direct control over how your voice AI Agents sound and what they cost: Generative for the most expressive, human-like speech when voice quality comes first, Neural for natural-sounding speech at a lower cost, and Standard for basic speech at the lowest cost.

Other Improvements

Cognigy

• Reduced NLU model loading time during intent scoring, helping AI Agents respond faster when many users interact with them simultaneously
• Renamed Salesforce Messaging for In-App and Web (MIAW) to Salesforce Enhanced Chat in the UI to match the current Salesforce product name
• Added the Custom URL field in the LLM editor for Mistral AI models
• Added the A2A Agent Node as a child Node of the LLM Prompt Node, making it easier to use A2A capabilities within LLM-based workflows
• Added API Key to the Performed By column and filter on the Audit Events page to make changes made through the Cognigy API easier to track
• Improved service security and stability

Cognigy Agent Copilot

• Added the End Session On Handover Close toggle and the Session TTL On Handover Close field to the Copilot section of the Endpoint settings. This improvement lets human agents trigger follow-up actions in the Agent Copilot Flow after resolving a conversation without keeping it open
• Added the Enable Branding toggle to the Copilot section of the Endpoint settings. This toggle lets you remove the Powered by Cognigy.AI branding to better align the Agent Copilot workspace with your organization’s branding

Cognigy Live Agent

• Improved service security, stability and performance

For further information, check out our complete Release Notes.