Which AI Models Power Zapier Automation? A Complete Guide
New AI models launch almost weekly, and keeping track of which one to use for your specific workflow feels like a full-time job. Think of this guide as your living reference for everything you need to know.
This article walks through the major AI providers available on Zapier, the models you can integrate into your Zaps right now, and what each model does best—all based on Zapier's AutomationBench testing.
How to Automate AI Models on Zapier
Zapier is the most broadly connected AI orchestration platform, integrating with thousands of apps from partners like Google, Salesforce, and Microsoft. Use forms, databases, and logic to build secure, AI-powered automation for critical workflows across your entire tech stack.
You have two ways to bring AI models into your automated workflows: use Zapier's built-in AI step (AI by Zapier) or connect directly to providers.
- AI by Zapier bundles dozens of models from OpenAI, Anthropic, Google, and others into a single step. No separate account needed, no token-based billing—you just pay per task executed on Zapier.
- Direct provider connections link straight to the provider's account. You trade simplicity for model variety—think older GPT versions or DeepSeek—and pay by token through your own API key.
There's also another option: if you want to trigger workflows directly from your AI assistant, use Zapier MCP. Install it into any app supporting the MCP standard and run actions across 9,000+ tools without leaving your chat window.
OpenAI Models (ChatGPT)
OpenAI's lineup is the broadest on Zapier—everything from budget-friendly mini models to advanced reasoning tools and specialized transcription and image generation options.
Best for: Sales and Marketing workflows. GPT-5.5 (XHigh) tops the leaderboard for both—the only model outside Anthropic's product line to lead a category. Check AutomationBench rankings.
What's new: OpenAI launched GPT-5.5 for the most complex professional work: coding, research, data analysis, and multi-step automation across multiple tools. In Zapier's early AutomationBench tests, GPT-5.5 became the first model to exceed 10% on the performance threshold.
| Model | Best For | Input | Output | Context Window | Output Price (per 1M tokens) |
|---|---|---|---|---|---|
| GPT-5.5 Pro | Problems requiring deep reasoning and maximum reliability, where accuracy matters more than speed or cost | Text, images | Text | 1M tokens | $180 |
| GPT-5.5 | Complex professional work: coding, research, data analysis, multi-step automation across tools | Text, images | Text | 1M tokens | $30 |
| GPT-5.4 Nano | High-volume, repetitive tasks where speed and cost matter most—classification, data extraction, ranking | Text, images | Text | 400K tokens | $1.25 |
| GPT-5.4 Mini | Complex, multi-step workflows requiring quick reasoning across content types and tools | Text, images | Text | 400K tokens | $4.50 |
| GPT-5.4 | Complex professional workflows needing rigorous logic and careful planning | Text, images, audio | Text | 1.05M tokens | $15 |
| GPT-5.3 | Quick, context-aware conversation and search | Text, images | Text | 128K tokens | $14 |
| GPT-5.2 | Advanced coding and automation with reliable multi-step reasoning | Text, images | Text | 128K tokens | $14 |
| GPT-5 Mini | Reasoning and logic for well-defined tasks | Text, images | Text | 400K tokens | $2 |
| GPT-5 Nano | Cost-efficient reasoning and logic for summaries, classification, and simple tasks | Text | Text | 400K tokens | $0.40 |
| GPT-4o Mini | Multimodal on a budget | Text, images, audio | Text | 128K tokens | $0.60 |
| GPT-4o | Multimodal tasks, especially voice and image—natural, human-like interaction | Text, images, audio, video | Text | 128K tokens | $10 |
| GPT-4.1 Mini | Balance between power, speed, and affordability for general-purpose work | Text, images | Text | 1.05M tokens | $1.60 |
| GPT-4.1 | Complex tasks that don't need advanced reasoning, with very long context windows | Text, images | Text | 1.05M tokens | $8 |
| GPT-4.1 Nano | Simple tasks where speed and cost beat raw power | Text | Text | 1.05M tokens | $0.40 |
| o4-mini | Fast, budget-friendly reasoning | Text, images | Text | 200K tokens | $4.40 |
| o3-mini | Lighter, cheaper alternative to o3 for reasoning-heavy tasks | Text | Text | 200K tokens | $4.40 |
| o3 | Advanced reasoning and logic | Text, images | Text | 200K tokens | $8 |
| GPT Image 1.5 | Latest image generation | Text, images | Text, images | N/A | $10 |
| GPT Image 1 | Image generation | Text, images | Images | N/A | $40 |
Anthropic Models (Claude)
Claude's models are known for excellent writing quality, careful instruction-following, and a design philosophy that prioritizes safety. Claude is a popular choice for tasks like long-form content creation, document analysis, and customer-facing chatbots that need natural, conversational tone.
Best for: Operations, Support, Finance, and HR workflows. Fable 5.0 (Max) leads Operations and HR, the newly released Sonnet 5 (Max) leads Finance, and Opus 4.8 (XHigh) leads Support. Opus 4.8 (Max) ranks second in Sales, and Sonnet 5 (Max) ties for second in Operations.
What's new: Sonnet 5 is now available on Zapier. It delivers Opus-level performance at Sonnet-tier pricing and works over twice as efficiently as Sonnet 4.6 on multi-step workflows. The biggest win is completion: Sonnet 5 trusts what it finds, understands how your apps work, and finishes tasks that would stall on Sonnet 4.6.
| Model | Best For | Input | Output | Context Window | Output Price (per 1M tokens) |
|---|---|---|---|---|---|
| Sonnet 5 | Coding, agents, and daily professional work at scale | Text, images | Text | 1M tokens | $15 |
| Fable 5.0 | Work requiring the highest reasoning and long-term vision | Text, images | Text | 1M tokens | $50 |
| Sonnet 4.6 | Coding, agents, business workflows—best price-to-performance balance | Text, images | Text | 1M tokens | $15 |
| Opus 4.8 | Complex reasoning, agentic coding, long-running tasks | Text, images | Text | 1M tokens | $25 |
| Opus 4.7 | Complex reasoning, agentic coding | Text, images | Text | 1M tokens | $25 |
| Opus 4.6 | Complex reasoning, coding, long-term tasks | Text, images | Text | 1M tokens | $25 |
| Haiku 4.5 | High-volume, latency-sensitive, budget-conscious tasks | Text, images | Text | 200K tokens | $5 |
| Sonnet 4.5 | Complex agents, coding; highest general intelligence | Text, images | Text | 200K tokens | $15 |
| Opus 4.1 | Complex reasoning, analysis, creative tasks | Text, images | Text | 200K tokens | $75 |
| Sonnet 4 | Balanced coding and workflows | Text, images | Text | 200K tokens | $15 |
| Haiku 3 | Quick, simple classification and cost-efficient work | Text, images | Text | 200K tokens | $1.25 |
Google Gemini Models
Google's Gemini lineup stands out for large context windows, competitive pricing, and strong multimodal capabilities across text, images, audio, and video. Gemini models excel at processing long documents, research-heavy workflows, and situations where keeping costs down matters.
Best for: High-volume workflows where cost is critical. Gemini 3.5 Flash (Medium) ranks fifth overall with 14.5%—just three points behind the top score, yet costs only around a quarter per task (0.87 USD vs. 3.67 USD). At that price, it's also the most affordable model on the list.
What's new: Gemini 3.5 Flash is now live on Zapier. It excels at orchestrating steps and following strict policies—the kind of work that trips up other models. That said, it can struggle with rigid output formats and decisions based on math it has to do itself.
| Model | Best For | Input | Output | Context Window | Output Price (per 1M tokens) |
|---|---|---|---|---|---|
| Gemini 3.5 Flash | Sub-agent deployment, multi-step workflows, and long-running tasks at scale | Text, images, audio, video, PDF | Text | 1M tokens | $9 |
| Gemini 3.1 Pro | Complex reasoning, demanding coding, massive data synthesis | Text, images, audio, video, PDF | Text, code, reasoning | 1M tokens | $30 |
| Gemini 3 Flash | High-speed automation, real-time chat, cost-conscious scaling | Text, images, audio, video | Text, code | 1M tokens | $0.30 |
| Gemini 3 Pro | Balance professional workflows with creative content | Text, images, audio, video | Text, code | 1M tokens | $3.75 |
| Gemini 2.0 Flash Lite | Basic classification, ultra-low latency tasks, simple extraction | Text, images | Text | 1M tokens | $0.15 |
| Gemini 2.0 Flash | Backward compatibility for high-performance 2.0-era applications | Text, images, audio | Text, code | 1M tokens | $0.30 |
| Gemini 2.5 Pro | Detailed multimodal analysis with high accuracy for legacy pipelines | Text, images, audio, video | Text, code | 2M tokens | $10.50 |
| Gemini 2.5 Flash | Production-level speed for multimodal processing | Text, images, audio, video | Text, code | 1M tokens | $0.90 |
| Stable Image Ultra Pro | High-quality professional image creation and editing | Text, images | Images | N/A | $0.05 |
Connect With Today's Best AI Models on Zapier
Whether you're just starting with AI automation or already building complex multi-step workflows, Zapier gives you the flexibility to choose the tools and models that actually fit your needs.
Description: Explore the best AI models available on Zapier—from OpenAI and Anthropic to Google Gemini. Find which model fits your workflow.
No Comment to " Which AI Models Power Zapier Automation? A Complete Guide "