AI Tools Reference

Use this when you're deciding which tool to try or comparing options at a glance. Free tiers, strengths, and best-for summaries in one place.

This reference covers the tools mentioned throughout the book. Pricing and features change frequently, so treat this as orientation, a starting point you confirm before you buy. Always verify current pricing and capabilities at each tool's website before subscribing. All prices are approximate USD as of 2026.

General-Purpose Chat AI

These are your primary tools and the foundation of the CLEAR system. Start with one from this category before exploring any other.

ChatGPT (OpenAI) Category: General-purpose chat AI. Free tier: Yes, with limits on files, images and the strongest models. Paid: around $20/month (Plus), from $100/month (Pro). Best for: Broad professional tasks, code, data analysis, image generation. Note: Most widely used; strong ecosystem of plugins and integrations

Claude (Anthropic) Category: General-purpose chat AI. Free tier: Yes, Claude Sonnet with limits. Paid: around $20/month (Pro). Best for: Long document analysis, nuanced writing, following complex instructions. Note: Handles very long documents in a single session; strong at tone and style

Gemini (Google) Category: General-purpose chat AI. Free tier: Yes, Gemini Flash. Paid: around $20/month (Google AI Pro). Best for: Research with web search, Google Workspace integration, multimodal tasks. Note: Deep integration with Google Docs, Sheets, and Gmail

Microsoft Copilot Category: General-purpose + Microsoft 365 integration. Free tier: Yes, limited. Paid: Copilot Chat is included with Microsoft 365; full Copilot in the apps is a paid add-on, around $21/user/month. Best for: Word, Excel, Teams, Outlook, AI embedded directly in Microsoft apps. Note: Best value if your organization already uses Microsoft 365

DeepSeek Category: General-purpose chat AI, and its thinking mode is also listed under Reasoning Models below. Free tier: Yes, full capability. Paid: Very low API pricing; free web interface. Best for: Strong reasoning and analytical capability, cost-sensitive users. Note: Chinese company; review your organization's data policies before use

Grok (xAI) Category: General-purpose + real-time information. Free tier: Yes, limited, on grok.com and through X. Paid: around $8/month (X Premium) to $30/month (SuperGrok standalone). Best for: Current events, trend monitoring, fast-moving topics. Note: Best for users who need real-time information access

Reasoning Models

Use these for complex multi-step analysis, planning, and tasks where the quality of reasoning matters more than speed. Slower and typically more expensive than standard models.

ChatGPT Thinking (OpenAI) Category: Reasoning model. Free tier: Limited, through the Thinking option. Paid: More use with ChatGPT Plus/Pro. Best for: Complex logic, multi-step planning, coding problems, financial analysis. Note: Noticeably slower than standard fast models; use when output quality justifies the wait

Claude (Thinking) Category: Reasoning model. Free tier: Limited; check the plan page. Paid: Included with Claude Pro on select models. Best for: Long-form analysis, research synthesis, complex document review. Note: Visible 'thinking' output shows the model's reasoning process

DeepSeek (thinking mode) Category: Reasoning model. Free tier: Yes, through the DeepSeek web interface. Paid: Very low API pricing. Best for: Heavy analytical reasoning on a budget, and a cheap second opinion on a hard problem. Note: Chinese company; review your organization's data policies before use

Image Generation

Midjourney Category: Image generation. Free tier: No. Paid: around $10/month. Best for: High-quality artistic and photographic images, design concept generation. Note: Runs in the browser and through Discord; aimed at stylised and photographic work rather than diagrams or text-heavy images

ChatGPT Images (OpenAI, formerly DALL-E) Category: Image generation. Free tier: Limited via ChatGPT free. Paid: Included with ChatGPT Plus. Best for: Quick concept visuals, integrated workflow with ChatGPT text tasks. Note: Most convenient if you already use ChatGPT

Adobe Firefly Category: Image generation. Free tier: Yes, limited credits. Paid: Included with Creative Cloud plans. Best for: Commercially licensed outputs, Photoshop integration. Note: Best for professionals who need legally safe generated images

Canva AI Category: Image generation + design. Free tier: Yes, limited features. Paid: around $15/month (Canva Pro). Best for: Non-designers creating presentations, social graphics, marketing materials. Note: Combines AI generation with pre-built templates; low barrier to entry

Audio, Music, and Voice

ElevenLabs Category: Voice synthesis and cloning. Free tier: Yes, limited characters/month. Paid: around $5/month. Best for: AI voiceovers, podcast narration, video narration. Note: Aimed at production narration rather than quick previews; voice cloning available on paid plans

Suno Category: Music generation. Free tier: Yes, 10 songs/day. Paid: around $10/month. Best for: Full song generation from text prompts, background music for video. Note: Produces complete songs with lyrics, melody, and production

Descript Category: Audio and video editing. Free tier: Yes, limited media minutes. Paid: around $24/month. Best for: Podcast editing, filler word removal, transcript-based editing. Note: Edit audio by editing the transcript; significantly reduces production time

Video Generation

Runway Category: Video generation and editing. Free tier: Yes, limited credits. Paid: around $12/month. Best for: AI video generation from text or images, professional video editing tools. Note: Text-to-video and image-to-video sit alongside a conventional editing timeline in one tool

HeyGen Category: AI avatar video. Free tier: Yes, limited. Paid: around $29/month. Best for: Talking-head videos, spokesperson content, multilingual video dubbing. Note: Translates video content into other languages while preserving lip sync

Code and Technical Tools

GitHub Copilot Category: Code completion and generation. Free tier: Yes, limited completions and chat. Paid: around $10/month (Copilot Pro). Best for: Real-time code suggestions, code explanation, debugging. Note: Integrates directly into VS Code and other editors; standard for developers

Cursor Category: AI-native code editor. Free tier: Yes, limited. Paid: around $20/month. Best for: Reasoning about entire codebases, complex refactoring, multi-file editing. Note: Faster than Copilot for larger tasks; preferred by developers working on full projects

Productivity and Research

Perplexity Category: AI-powered research. Free tier: Yes, limited. Paid: around $20/month (Pro). Best for: Research with cited sources, current information, topic synthesis. Note: Provides source citations with every response; strong alternative to search

Notion AI Category: Writing and knowledge management. Free tier: A short AI trial. Paid: Included with Notion Business ($20/user/month billed annually). Best for: Summarizing documents, generating content inside Notion, organizing notes. Note: AI comes bundled with the Business and Enterprise plans rather than sold as a separate add-on. Notion has repackaged this before, so check the current plan page

Zapier (AI features) Category: Workflow automation. Free tier: Yes, limited automations. Paid: around $20/month. Best for: Connecting AI tools to automate workflows between apps. Note: Allows AI-triggered automations across hundreds of business tools

Model Types: Fast vs. Reasoning

Every general-purpose AI platform now offers at least two types of models. Knowing when to use each is one of the highest-leverage decisions in your AI practice.

Fast Models (Default Models)

Examples: ChatGPT's default model, Claude Sonnet, Gemini Flash

What they do: Process your prompt and generate a response immediately, without showing their work. Fast, fluid, and fluent.

Best for: Daily drafts, emails, summaries, brainstorming, explanations, research synthesis, creative variations, anything where speed matters and the task is familiar.

Cost: Usually included in free and standard paid tiers.

Use by default: Start here for everything. The vast majority of professional work is well-served by fast models.

Reasoning Models

Examples: ChatGPT Thinking, Claude with Thinking on, Gemini's Thinking levels

What they do: Pause before responding. Work through the problem step by step, showing their reasoning before delivering an answer. Slower, more deliberate, visibly transparent about uncertainty.

Best for: Complex logic problems, multi-step planning, code debugging, financial or legal analysis, tasks where intermediate reasoning steps matter as much as the final answer.

Cost: Typically require paid tiers; some available with limits on free plans.

Use when: Accuracy is critical and speed is secondary. The extra wait is worth it when getting it right the first time saves rework.

Chain-of-Thought Prompting: Getting Reasoning Behaviour Without a Model Switch

Add any of these phrases to any standard prompt to activate more deliberate, step-by-step thinking from your fast model:

"Think step by step." | "Show your reasoning before giving your answer." | "List your assumptions first." | "Walk me through how you arrived at this."

These phrases trigger Chain-of-Thought prompting: the model works through the problem methodically before responding, improving accuracy on complex tasks without the cost or wait of a dedicated reasoning model.

Search-Enabled Mode

Use when your task requires information after the model's training cutoff: recent news, updated statistics, current events, or any fact that could have changed.

Gemini: Web search built in by default. Perplexity: Search is the core function, all responses include cited sources. ChatGPT: Search often activates automatically; can be toggled manually. Claude: Web search on every plan, free included. DeepSeek: Click the search toggle before submitting your prompt.

Building Your AI Stack

The Starter Stack (Free)

Claude: Long documents, nuanced writing, complex analysis.

ChatGPT: Brainstorming, quick drafts, creative variations.

Gemini: Research with live web data; Google Workspace integration.

DeepSeek: Heavy analytical reasoning when you need a second opinion.

For most professionals, this free combination covers the daily work. Start here. Add paid subscriptions only when a specific tool becomes a consistent bottleneck.

The Upgrade Decision

Are you hitting free tier limits regularly? (Rate limits, capacity messages, forced waits)

Does AI save you two or more hours per week on tasks you previously did manually?

Do you need advanced features? (Extended context, reasoning models, image generation, file analysis)

If two or more of these are true, upgrading one tool's paid tier is likely worth it. Upgrade the one that's creating the bottleneck. Leave everything else on free.

Upgrade priority for most professionals:

First upgrade: Claude Pro or ChatGPT Plus, whichever tool you use most. $20/month each.

Second upgrade: Specialist tool that fits your workflow (ElevenLabs for voice, Midjourney for images, etc.).

Microsoft users: Copilot with Microsoft 365 may make both unnecessary if your work runs entirely in that ecosystem.