AI Prompt (Multi-LLM)

This action is suited to execute LLM-prompts from multiple AI Providers, including on-premises ones based on Ollama.

Supported LLM Providers and Models

Cohere

• command-r
• command-r-plus
• gemini-pro

Anthropic

• claude-3-5-sonnet

Groq

Grouped by “creator”

Meta (LLaMA)

• llama-3.2-11b-vision-preview
• llama-3.1-8b-instant
• llama-3.2-3b-preview
• llama-3.2-90b-vision-preview
• llama-guard-3-8b
• llama3-8b-8192
• llama-3.2-1b-preview
• llama-3.3-70b-versatile
• llama3-70b-8192
• llama-3.3-70b-specdec

DeepSeek / Alibaba Cloud

• deepseek-r1-distill-qwen-32b
• deepseek-r1-distill-llama-70b

Alibaba Cloud

• qwen-2.5-32b
• gemma2-9b-it

SDAIA

• allam-2-7b

Mistral AI

• mixtral-8x7b-32768

Ollama

• User defined models

How to Configure and Use the Action

Drag & Drop the Action from the actions palette on the left to the stage:

Once on the stage you can configure your favourite LLM prompt using the action detail panel.

To properly use your Action you need to configure it in the Settings > Integrations section, providing the corresponding LLM API-KEY.

For example, for Google Gemini:

Using Ollama with AI Prompt Action

As with other LLMs, in order to use Ollama with the AI Prompt action, you need to configure the Ollama Integration on Pingstreams by going to Settings → Integrations, entering:

• The URL of the machine where Ollama is running
• (Optional) Your favorite models to use for faster action configuration

To add a model to your Favorites List, type the exact model name and press the Enter button. Finally save the settings clicking on Save button.

In the AI Prompt action, select Ollama as LLM and choose a model between the predefined favorite models

Choosing the Right LLM for Your Use Case

Different LLMs excel at different tasks. Here’s a guide to help you choose:

For General Conversations & Customer Support

• Claude 3.5 Sonnet: Excellent for nuanced conversations and customer service
• GPT-4o: Great all-around model for most conversational tasks
• Llama 3.1-8b: Good balance of performance and speed for general use

For Creative Content Generation

• GPT-4o: Excellent for creative writing and content generation
• Claude 3.5 Sonnet: Strong creative abilities with good reasoning
• Llama 3.3-70b: Good for creative tasks requiring more context

For Technical Support & Code

• GPT-4o: Excellent for technical explanations and code generation
• Claude 3.5 Sonnet: Strong analytical abilities for technical content
• DeepSeek models: Specialized for coding and technical tasks

For Multi-language Support

• GPT-4o: Strong multi-language capabilities
• Qwen 2.5-32b: Excellent for Chinese and other Asian languages
• Gemini Pro: Good multi-language support

For High-Volume, Fast Responses

• Llama 3.1-8b-instant: Optimized for speed
• Groq models: Generally faster inference times
• Local Ollama models: No API limits, full control

For Privacy & On-Premises

• Ollama models: Run locally, complete data privacy
• Custom fine-tuned models: Via Ollama for specific use cases

Best Practices

1. Test Different Models: Start with GPT-4o or Claude 3.5 Sonnet, then experiment
2. Consider Response Time: Lighter models respond faster but may be less capable
3. Monitor Costs: Larger models cost more per token
4. Use Context Wisely: Include relevant conversation history in your prompts
5. Set Appropriate Token Limits: Balance response quality with cost and speed

We hope you enjoy our Multi-LLM Action that will let you use your favourite LLM provider and models!

If you have questions about the AI Prompt Action or other Pingstreams features feel free to send us an email to support@pingstreams.io or leave us feedback.