Agent Platform supports a wide range of AI models—platform-hosted, open-source, externally integrated, and third-party. Choose based on your use case: agentic workflows, real-time voice, or background processing.
Models for Agentic Apps and Agents
External Models
Agentic Apps support Agent and Supervisor orchestration with models from these providers:
Open-Source Models
Agentic Apps support select open-source models from these providers:
| Provider | Model Variants |
|---|
| Meta-llama | - meta-llama/Meta-Llama-3.1-8B-Instruct
- meta-llama/Llama-3.2-1B-Instruct
- meta-llama/Llama-3.2-3B-Instruct
|
| Mistralai | - mistralai/Mistral-7B-Instruct-v0.3
- mistralai/Mistral-Nemo-Instruct-2407
|
| Xiaomimimo | |
Real-Time Voice Models
Real-Time Voice is available only with the following models:
| Provider | Model Variants |
|---|
| OpenAI | - gpt-4o-realtime-preview
- gpt-4o-mini-realtime-preview
|
| Google Gemini | - gemini-live-2.5-flash-preview
- gemini-2.0-flash-live-001
|
| Azure OpenAI | - GPT-Realtime
- GPT-Realtime-Mini
|
Custom Model Integration
To use a custom third-party model in Agentic Apps, it must meet two requirements:
Tool Calling Support (Mandatory)
The model must natively support Tool Calling. Models without this capability cannot be used in Agentic Apps.
If supported, enable Tool Calling explicitly during model configuration.
Learn how to enable model features
Compatible API Structure
The request and response structure must follow either:
- Anthropic – Messages API
- OpenAI – Chat Completions API
For setup steps, see Add an External Model Using API Integration.
Platform-Hosted Open-Source Models
The platform hosts 30+ curated open-source models, available as a service and optimizable before deployment.
| Provider | Model Variants |
|---|
| Argilla | - argilla/notus-7b-v1
- argilla/notux-8x7b-v1
|
| DeepSeek | - deepseek-ai/DeepSeek-R1-Distill-Llama-8B
- deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B
- deepseek-ai/DeepSeek-R1-Distill-Qwen-14B
- deepseek-ai/DeepSeek-R1-Distill-Qwen-7B
|
| EleutherAI | - EleutherAI/gpt-j-6b
- EleutherAI/gpt-neo-1.3B
- EleutherAI/gpt-neo-125m
- EleutherAI/gpt-neo-2.7B
- EleutherAI/gpt-neox-20b
|
| Facebook | - facebook/opt-1.3b
- facebook/opt-2.7b
- facebook/opt-350m
- facebook/opt-6.7b
|
| Google | - google/flan-t5-base
- google/flan-t5-large
- google/flan-t5-small
- google/flan-t5-xl
- google/flan-t5-xxl
- google/gemma-2-27b-it
- google/gemma-2-9b-it
- google/gemma-2b
- google/gemma-2b-it
- google/gemma-3-12b-it
- google/gemma-7b
- google/gemma-7b-it
|
| Helsinki-NLP | - Helsinki-NLP/opus-mt-es-en
|
| HuggingFaceH4 | - HuggingFaceH4/zephyr-7b-alpha
- HuggingFaceH4/zephyr-7b-beta
|
| Meta-llama | - meta-llama/Llama-2-13b-hf
- meta-llama/Llama-2-7b-hf
- meta-llama/Llama-3.2-1B
- meta-llama/Llama-3.2-1B-Instruct
- meta-llama/Llama-3.2-3B
- meta-llama/Llama-3.2-3B-Instruct
- meta-llama/Llama-3.2-11B-Vision-Instruct
- meta-llama/Llama-Guard-4-12B
- meta-llama/Meta-Llama-3-8B
- meta-llama/Meta-Llama-3-8B-Instruct
- meta-llama/Meta-Llama-3.1-8B
- meta-llama/Meta-Llama-3.1-8B-Instruct
|
| Microsoft | - microsoft/phi-1
- microsoft/phi-1_5
- microsoft/phi-2
- microsoft/Phi-3-medium-128k-instruct
- microsoft/Phi-3-medium-4k-instruct
- microsoft/Phi-3-mini-128k-instruct
- microsoft/Phi-3-mini-4k-instruct
|
| Mistralai | - mistralai/Mistral-7B-Instruct-v0.1
- mistralai/Mistral-7B-Instruct-v0.2
- mistralai/Mistral-7B-Instruct-v0.3
- mistralai/Mistral-7B-v0.1
- mistralai/Mistral-Nemo-Instruct-2407
- mistralai/Mixtral-8x7B-Instruct-v0.1
- mistralai/Mixtral-8x7B-v0.1
|
| OpenAI | |
| OpenAI Community | - openai-community/gpt2-large
- openai-community/gpt2-medium
- openai-community/gpt2-xl
|
| Stable Diffusion | - stabilityai/stable-diffusion-xl-base-1.0
- stabilityai/stable-diffusion-2-1
- stable-diffusion-v1-5/stable-diffusion-v1-5 (Available only in the text-to-image node; no Prompt Studio support)
|
| T5 | |
| Tiiuae | - tiiuae/falcon-40b
- tiiuae/falcon-40b-instruct
- tiiuae/falcon-7b
- tiiuae/falcon-7b-instruct
- tiiuae/falcon-rw-1b
|
| Xiaomimimo | |
Structured Output
Platform-hosted models can produce structured JSON responses, making outputs consistent and easy to parse.
Supported optimization techniques: No optimization or vLLM only.
CT2-optimized, fine-tuned, Hugging Face imported, and locally imported models do not support structured output.
| Model | vLLM | No Optimization |
|---|
| amazon/MistralLite | ❌ | ✅ |
| argilla/notus-7b-v1 | ❌ | ✅ |
| EleutherAI/gpt-j-6b | ❌ | ✅ |
| facebook/opt-1.3b | ✅ | ✅ |
| facebook/opt-2.7b | ✅ | ✅ |
| facebook/opt-350m | ✅ | ✅ |
| facebook/opt-6.7b | ✅ | ✅ |
| google/gemma-2b | ❌ | ✅ |
| google/gemma-2b-it | ❌ | ✅ |
| google/gemma-7b | ❌ | ✅ |
| google/gemma-7b-it | ❌ | ✅ |
| HuggingFaceH4/zephyr-7b-alpha | ❌ | ✅ |
| HuggingFaceH4/zephyr-7b-beta | ❌ | ✅ |
| meta-llama/Llama-2-7b-chat-hf | ❌ | ❌ |
| meta-llama/Llama-2-7b-hf | ❌ | ✅ |
| meta-llama/Llama-3.2-1B | ✅ | ❌ |
| meta-llama/Llama-3.2-1B-Instruct | ✅ | ❌ |
| meta-llama/Llama-3.2-3B | ✅ | ❌ |
| meta-llama/Llama-3.2-3B-Instruct | ✅ | ❌ |
| meta-llama/Meta-Llama-3-8B | ✅ | ✅ |
| meta-llama/Meta-Llama-3-8B-Instruct | ❌ | ✅ |
| meta-llama/Meta-Llama-3.1-8B | ✅ | ✅ |
| meta-llama/Meta-Llama-3.1-8B-Instruct | ❌ | ✅ |
| microsoft/Phi-3-medium-128k-instruct | ❌ | ✅ |
| microsoft/Phi-3-medium-4k-instruct | ❌ | ✅ |
| microsoft/Phi-3-mini-128k-instruct | ✅ | ✅ |
| microsoft/Phi-3-mini-4k-instruct | ❌ | ✅ |
| microsoft/phi-1 | ❌ | ✅ |
| microsoft/phi-1_5 | ❌ | ✅ |
| microsoft/phi-2 | ✅ | ✅ |
| mistralai/Mistral-7B-Instruct-v0.1 | ❌ | ✅ |
| mistralai/Mistral-7B-Instruct-v0.2 | ❌ | ✅ |
| mistralai/Mistral-7B-Instruct-v0.3 | ✅ | ✅ |
| mistralai/Mistral-7B-v0.1 | ❌ | ✅ |
| openai-community/gpt2-large | ❌ | ✅ |
| openai-community/gpt2-medium | ❌ | ✅ |
| openai-community/gpt2-xl | ❌ | ✅ |
| tiiuae/falcon-7b | ❌ | ✅ |
| tiiuae/falcon-7b-instruct | ❌ | ✅ |
| tiiuae/falcon-rw-1b | ✅ | ✅ |
External Models for Easy Integration
With Easy Integration, connect to external model providers without any infrastructure setup—just authenticate and start using models in flows, tools, or agents.
