Convoflow integrates with leading AI model providers. This enables the Agent node and reasoning blocks to understand natural language, follow complex business guidelines, and execute visual tool subflows autonomously.
Available models in the Agent node
When configuring the Agent node on your flow canvas, you can choose the provider and model to process conversation turns. The following models are supported:
| Provider | Available models | Focus and recommended use cases |
|---|---|---|
| OpenRouter | • Claude Sonnet 4 • Claude Haiku 4.5 • DeepSeek Chat • GLM 5.2 • GLM 5.3 Flash • GLM 5.3 | High multimodel versatility, deep reasoning with Claude, great tool orchestration, and fast, cost-effective options with GLM and DeepSeek. |
| OpenAI | • GPT 5.4 • GPT 5.4 Mini • GPT 5.4 Nano • GPT 5.6 Luna • GPT 5.6 Terra • GPT 5.6 Sol | Industry-leading instruction following, autonomous agent workflows, and cost-effective speed options (Mini and Nano) or multimodal reasoning (GPT 5.6 Luna, Terra, and Sol). |
| DeepSeek | • DeepSeek V4 Flash • DeepSeek V4 Pro | Strong technical reasoning and broad contextual understanding with rapid turnaround (Flash) or deep analysis (Pro). |
| Groq | • GPT OSS 120B • GPT OSS 20B | Ultra-low latency inference on specialized hardware (LPU), ideal for real-time conversations where instantaneous responsiveness is required. |
Selecting a model in your flow
- On the visual canvas, open the property panel for the Agent node.
- In the Provider (
modelProvider) field, choose your desired platform (openrouter,openai,deepseek, orgroq). - In the Model (
modelId) field, select the model fitting your task complexity and speed requirements. - Wire any required tools using subflows so the agent can look up data or trigger external actions before answering.