Product Reference · Updated October 13, 2026
One Multi-Model AI Client for ChatGPT, Claude, Gemini, DeepSeek and More
tt brings OpenAI, Anthropic, Google, DeepSeek and other model providers and managed AI platforms into one native BYOK client, so you can connect the services you use without moving conversations between separate apps.
tt is a bring-your-own-key client. You connect services available to your own accounts. Standard chat and Team Mode use every service whose model and credential configuration is complete; configure one service for a single-model chat or several for a multi-model chat. tt does not bundle an AI subscription or resell provider usage.
Model Providers
tt provides connection configurations for major model providers. Exact model choices can change as providers update their APIs, so the current in-app configuration list is the source of truth.
Available provider integrations
- OpenAI (GPT / ChatGPT)
- Anthropic (Claude)
- Google (Gemini)
- xAI (Grok)
- DeepSeek
- VolcEngine Ark (Doubao)
- Qwen
- Moonshot AI (Kimi)
- Zhipu AI (GLM)
- MiniMax
Managed AI Platforms
If your models are accessed through a cloud or enterprise platform, tt also provides integrations for provider-specific credentials, regions, and endpoints where required.
Available hosting-platform integrations
- AWS Bedrock
- Azure AI
- Google Vertex AI
- Alibaba Cloud Model Studio
What Integration Support Means
Your Account, Your Credentials
You need an eligible account and valid credentials from each service. Provider billing, quotas, region access, and content rules still apply.
Current Models Stay Flexible
Model names and availability evolve. tt maintains provider connections while its configuration list can be updated over time.
One Place to Work
Once configured, services appear together in tt. Every service with a complete model and credential configuration participates in the chat workflow.
Original Responses Remain Visible
In a multi-model standard chat, you can inspect each completed response as well as the synthesized answer.
Two Ways to Use Multiple Models
Standard multi-model chat
Send one question to all completely configured services, compare their original replies, and let your chosen summarizer organize the result into one answer. Use the repeatable AI-answer comparison method when you need a more rigorous evaluation.
Team Mode discussion
Give all completely configured services a shared conversation where they can react to one another, contribute new information, and stop responding when they have nothing new to add. See how Team Mode works.