Skip to main content

AI Platforms and Models

Written by Peter Novosel

When creating powerful automations with SmartSuite’s Custom Prompt action, selecting the right AI model is essential for delivering fast, reliable, and cost-effective results. To make this process easier, SmartSuite now presents a curated list of supported AI platforms and models, designed to help users choose the best model for their specific use case. Instead of exposing a raw, confusing list from LiteLLM—which often includes unstable or preview-only models—we’ve added a clean, informative interface that highlights key model characteristics with consistency and clarity.

Each model in the list is presented in a grid format organized by provider (such as OpenAI, Gemini, Anthropic, Perplexity, xAI, and IBM). The grid provides the following information:

  • Model Name: The official identifier of the model you can select.

  • Description: A short, user-friendly summary explaining what the model is best suited for—whether that’s summarization, reasoning, image analysis, or fast response generation.

  • Price Indicator: A relative cost indicator using dollar signs ($ to $$$$) to reflect general affordability. A single dollar sign represents the most cost-effective models, while four dollar signs represent premium options.

  • Speed Indicator: A scale from 1 to 3, with 3 being the fastest-performing models. This helps users choose models that match their latency expectations.

  • PDF & Image Scanning Support: Icons show whether a model supports input analysis of PDFs or images. Tooltips clarify this further when hovering over the icons (e.g., “This model can scan images”).

This structured approach empowers users to confidently select models based on the needs of their automation, whether they require multimodal support, budget-conscious processing, or maximum performance. The curated list also supports future flexibility, allowing SmartSuite to quickly add, remove, or update model information as AI platforms evolve.

OPEN AI

Model

Short Description

Price

Speed

PDFs

Images

GPT-4.1 Mini

Affordable GPT model for straightforward tasks.

$

3 (fastest)

❌

❌

GPT-4.1

Excellent for summarization and data extraction.

$$

2

❌

❌

o4-mini

Reasoning-focused, good for complex tasks.

$$$

2

✅

✅

o3-pro

Most powerful multimodal reasoning.

$$$$

1 (slowest)

✅

✅

o3

Multimodal reasoning for agentic tasks.

$$

2

✅

✅

GPT-4.1 Mini (with search)

GPT-4.1 Mini enhanced with web search.

$

3

❌

❌

GPT-4.1 (with search)

GPT-4.1 enhanced with built-in web search.

$$

2

❌

❌

GEMINI

Model

Description

Price

Speed

Scans PDFs

Images

gemini-2.5-flash

Fast, cost-efficient model with 1M‑token context. Great for summarization and image/audio/text tasks.

$

3

✅

✅

gemini-2.5-pro

High-end reasoning/coding powerhouse, full multimodal and Deep Think mode for challenging tasks.

$$$$

1

✅

✅

gemini-2.0-flash

Balanced stable model with multimodal input and quick responses.

$$

2

✅

✅

gemini-2.0-flash-lite

Lightweight, fastest and cheapest 2.0 option for high-throughput tasks.

$

3

✅

✅

ANTHROPIC

Model

Description

Price

Speed

Scan PDFs

Images

Claude 3.5 Haiku

Fastest Claude model, ideal for quick, lightweight tasks.

$$

3

✅

✅

Claude 3 Opus

Early premium model, good for complex reasoning.

$$$$

1

✅

✅

Claude 3.5 Sonnet

Balanced performance and cost, great for tool use.

$$$

2

✅

✅

Claude 3.7 Sonnet

Enhanced version of 3.5 Sonnet with better reasoning.

$$$

2

✅

✅

Claude 4 Sonnet

Multimodal mid-tier 4.x model, solid for content + tools.

$$$

2

✅

✅

Claude 4 Opus

Most capable Anthropic model for deep reasoning + tools.

$$$

1

✅

✅

PERPLEXITY

Model

Description

Price

Speed

Scan PDFs

Images

Sonar

Lightweight, fast model for quick, grounded search results.

$

3

✅

✅

Sonar Pro

Smarter search with deeper understanding and structured outputs.

$$

2

✅

✅

Sonar Reasoning

Tool-like model with multi-step reasoning and live search capabilities.

$$$

1

✅

✅

Sonar Reasoning Pro

Enhanced reasoning model for complex workflows and long-form queries.

$$$$

1

✅

✅

R1-1776

Specialized model for structured research and formal documentation tasks.

$$$

2

✅

✅

Sonar Deep Research

Expert-level research model using iterative search, citation, and synthesis over 2–4 minutes.

$$$

1

✅

✅

xAI

Model

Description

Price

Speed

Scan PDFs

Images

Grok-2-Vision-1212

8K context multimodal vision model (charts, screenshots, etc.)

$$

2

✅

✅

Grok-3-Beta

Flagship 131K token text-only Grok model

$$$

2

❌

❌

Grok-3-Mini-Beta

Fast, light version of Grok-3 for affordable inference

$

3

❌

❌

IBM

Model

Description

Price

Speed

PDFs

Images

Granite-3-2B

Lightweight instruct model with vision input

$$

2

✅

✅

Granite-3-8B

Mid-tier general-purpose multimodal LLM

$$

2

✅

✅

Llama-3-70B Vision

Vision-capable Llama model via IBM API

$$$

1–2

✅

✅

Related Articles
Did this answer your question?