| Model | Core Features & Limitations | Best Use Cases & Tips | Example Prompt |
|---|---|---|---|
|
GPT‑4.5 🧠 🧠 ❤️ ❤️+ Enhanced Conversational & Analytical Exels in creative writing, exploring new ideas and humanlike fluid conversation. |
⚡ Fast Processing 🧑💻 Sence of humor 🧑💻 Capable to reason with Gen Z 🔍 Integrated search Limitations: Context window 128 000 Otput tokens limit 16,384 May require fact‑checking |
Complex queries with multiple steps Creative writing and detailed analysis Tasks requiring balanced speed and depth Tips: ✓ Ideal for multi‑step inquiries ✓ Balance creative input with logical structuring |
✏️ In‑depth planning "Develop a strategy for a multi‑stage marketing campaign using emerging AI tools". 🦕 Style "The story in the style of David Sedaris, Patricia Lockwood, Simon Rich, and Lindy West. It's about two dogs who are trying to start an smm agency in Paris, during the economy crysis and the raise of generative AI" 💻 Methodology |
|
GPT‑4o 🧠 Conversational
Slightly deeper analysis than GPT‑4o mini |
⚡⚡ Fast Response 🗣️ Voice Enabled 🎨 DALL‑E Integration 💾 Contextual Memory Limitations: Input Context window 128 000 Otput tokens limit 16,384 Oct 01, 2023 knowledge cutoff Fact-check recommended |
Daily tasks Voice interactions Visual content generation Contextual chat or conversation Tips: ✓ Leverage multimodal features ✓ Structure prompts for context ✓ Verify outputs |
Compare Alice's size at the beginning vs. the end of the story 🗣️ Voice Q&A 🎨 Quick image prompt ✏️ Summaries with context |
|
o3‑mini + Advanced Reasoning
Multi‑step logic with balanced speed |
⚡⚡ Rapid Processing 💻 Advanced Code & Debugging 🔍 Contextual Search Optimized for technical tasks Limitations: 150 msgs/day (Plus) 8,192 tokens Complex queries suitable |
Coding & technical tasks Detailed analysis Research queries Multi‑step queries Tips: ✓ Combine search & coding ✓ Provide code examples ✓ Balanced approach for logic & speed |
Examine the Mad Hatter’s clock vs. his perpetual tea party 🧩 Solve puzzles 📝 Summarize text 💻 Generate code |
|
o3‑mini‑high ++ Reasoning
Extended chain‑of‑thought, really good at coding. Quick thinker |
💻 Expert Code Generation ⚡ Slower, Detailed Output 📄 Advanced Docs Analysis 🎯 High Accuracy Limitations: 50 requests/session 8,192 tokens Use for critical tasks |
Complex coding challenges Document analysis Logical problem‑solving High‑priority only Tips: ✓ Provide detailed input ✓ Accuracy is paramount ✓ Allow extra time |
Analyze paradoxes in Alice–Caterpillar dialogue 🧠 Deep logic 📄 Document dissection 🗂️ Cross‑reference arguments |
|
o1 Pro Mode +++ Supreme Reasoning
Highest expertise, though performance isn't different much from o3‑high‑mini |
💻 Superior Code 🎯 Max Accuracy & Knowledge Limitations: Pro tier only No token limits Precision over speed |
Critical analysis Complex problem‑solving Expert tasks (math/law) Errors unacceptable Tips: ✓ High‑stakes tasks only ✓ Provide domain context ✓ Allow time for thorough analysis |
Relate paradoxes in Alice–Caterpillar to set theory 🏆 Expert logic 🔬 In‑depth critique 💡 Theoretical expansions |
|
Deep Research ++++ Deep Analytical
Extensive data correlation, Great internet surfing skills |
🔍 Advanced Search 📊 Comprehensive Analysis 📈 Dynamic Visualization Limitations: Pro tier only Large-scale inputs, resource heavy |
Research projects Large data analysis Complex reports Academic/market research Tips: ✓ Structure queries (intro, data, conclusion) ✓ Verify with external sources ✓ Perfect for in‑depth reports ✓ If you’re feeling savvy, consider a “priming” stage • Set the context • Introduce specific details • Provide background |
I am going to publish Alice in Wonderland with my own illustrations. Research reliable printing companies and calculate the costs of production. 🔍 Data scanning 🔍 Web research 📊 Graph summary 📈 Trend analysis |
| Model | Core Features & Limitations | Best Use Cases & Tips | Example Prompt |
|---|---|---|---|
|
GPT‑4.5 🧠 🧠 ❤️ ❤️+ Enhanced Conversational & Analytical Excels in creative writing, exploring new ideas and humanlike fluid conversation. |
⚡ Fast Processing 🧑💻 Sense of humor 🧑💻 Capable to reason with Gen Z 🔍 Integrated search Limitations: Context window 128 000 Output tokens limit 16,384 May require fact‑checking |
Complex queries with multiple steps Creative writing and detailed analysis Tasks requiring balanced speed and depth Tips: ✓ Ideal for multi‑step inquiries ✓ Balance creative input with logical structuring |
✏️ In‑depth planning "Develop a multi‑phase narrative where Alice ventures through a reimagined Wonderland that mirrors our modern digital landscape using AI-inspired twists." 🦕 Style "Adopt a tone blending Lewis Carroll’s playful absurdity with a modern, witty commentary reminiscent of David Sedaris." 💻 Methodology "Structure the journey into distinct episodes, each reflecting evolving challenges and surreal solutions." |
|
GPT‑4o 🧠 Conversational
Slightly deeper analysis than GPT‑4o mini |
⚡⚡ Fast Response 🗣️ Voice Enabled 🎨 DALL‑E Integration 💾 Contextual Memory Limitations: Input Context window 128 000 Output tokens limit 16,384 Oct 01, 2023 knowledge cutoff Fact-check recommended |
Daily tasks Voice interactions Visual content generation Contextual chat or conversation Tips: ✓ Leverage multimodal features ✓ Structure prompts for context ✓ Verify outputs |
Compare Alice's size at the beginning vs. the end of the story 🗣️ Voice Q&A 🎨 Quick image prompt ✏️ Summaries with context |
|
o3‑mini + Advanced Reasoning
Multi‑step logic with balanced speed |
⚡⚡ Rapid Processing 💻 Advanced Code & Debugging 🔍 Contextual Search Optimized for technical tasks Limitations: 150 msgs/day (Plus) 8,192 tokens Complex queries suitable |
Coding & technical tasks Detailed analysis Research queries Multi‑step queries Tips: ✓ Combine search & coding ✓ Provide code examples ✓ Balanced approach for logic & speed |
Examine the Mad Hatter’s clock vs. his perpetual tea party 🧩 Solve puzzles 📝 Summarize text 💻 Generate code |
|
o3‑mini‑high ++ Reasoning
Extended chain‑of‑thought, really good at coding. Quick thinker |
💻 Expert Code Generation ⚡ Slower, Detailed Output 📄 Advanced Docs Analysis 🎯 High Accuracy Limitations: 50 requests/session 8,192 tokens Use for critical tasks |
Complex coding challenges Document analysis Logical problem‑solving High‑priority only Tips: ✓ Provide detailed input ✓ Accuracy is paramount ✓ Allow extra time |
Analyze paradoxes in Alice–Caterpillar dialogue 🧠 Deep logic 📄 Document dissection 🗂️ Cross‑reference arguments |
|
o1 Pro Mode +++ Supreme Reasoning
Highest expertise, though performance isn't different much from o3‑high‑mini |
💻 Superior Code 🎯 Max Accuracy & Knowledge Limitations: Pro tier only No token limits Precision over speed |
Critical analysis Complex problem‑solving Expert tasks (math/law) Errors unacceptable Tips: ✓ High‑stakes tasks only ✓ Provide domain context ✓ Allow time for thorough analysis |
Relate paradoxes in Alice–Caterpillar to set theory 🏆 Expert logic 🔬 In‑depth critique 💡 Theoretical expansions |
|
Deep Research ++++ Deep Analytical
Extensive data correlation, Great internet surfing skills |
🔍 Advanced Search 📊 Comprehensive Analysis 📈 Dynamic Visualization Limitations: Pro tier only Large-scale inputs, resource heavy |
Research projects Large data analysis Complex reports Academic/market research Tips: ✓ Structure queries (intro, data, conclusion) ✓ Verify with external sources ✓ Perfect for in‑depth reports ✓ If you’re feeling savvy, consider a “priming” stage • Set the context • Introduce specific details • Provide background |
I am going to publish Alice in Wonderland with my own illustrations. Research reliable printing companies and calculate the costs of production. 🔍 Data scanning 🔍 Web research 📊 Graph summary 📈 Trend analysis |
| Model name | Core Features & Limitations | Best Use Cases & Tips | Example Prompt |
|---|---|---|---|
|
Claude 3.7 Sonnet 🧠🧠 ++ Hybrid: Conversational with Advanced Reasoning A decent writer which is also very good at coding Update - March 2025 |
🎯 Strong Accuracy 💡 Complex Problem Solving 🖼️ Image Analysis 📊 Data Analysis Data Visualisation Limitations: 200K Input(150 000 words) 126K Output |
✓ Complex analysis; ✓ Creative writing; ✓ Coding; ✓ Problem solving; ✓ Data visualisation ✓ Text Stylisation Tips: Start with one-shot (detailed prompt); ...Ask for adjustments Try Extended Thinking Mode |
📊 Data Analysis “Analyze the following dataset to identify key trends and anomalies: [insert dataset].” 🧮 Problem Solving “Develop a step-by-step solution for this advanced calculus problem: [insert problem].” ✍️ Creative Content Generation “Compose a short story about a very lonely plant in the style of John Seabrook and Michael Cunningham.” 💻 Coding Assistance “Write a Python script that uses PyTorch for semantic segmentation of the image, with the ability to distinguish between 10 common household objects” ⏳ Extended Thinking Mode “With extended thinking enabled, provide a detailed analysis of the economic implications of renewable energy adoption.” |
|
Claude 3 Opus 🧠 ++ Expert Writing and Reasoning Big context window For multi-chapter summarisation Long formats Last Update - 2023 |
🎯 Creative and good with languages 🔬 Deep understanding of unstructured data 📚 Advanced Writing and conclusions 🖼️ Image Understanding Limitations: Large Input 200,000 tokens (±150 000 words) | Limited output - 4096 tokens. It can process a large book, but has a limit on the length of its responses. |
✓The reader’s companion ✓Effectively tracks complex ideas in large texts ✓Suggests unexpected yet high-quality metaphors and explanations—even if they are not explicitly stated ✓Expert-level tasks in the humanities ✓Performs well with translations |
📚 Text analysis. “Read this book and write a review on how AI is being used in neuroscience research” 📝 Creative writing “Propose an alternative ending to a classic novel, then analyze how this change would affect the main characters’ motivations and the story’s themes.” 🔍 Deep and creative insights “Here is a historical dataset with trends in global energy use. Identify patterns, propose a new hypothesis, and explain how you’d test it.” |
|
Claude 3 Haiku 🧠 – Quick Reasoning Optimized for conversational chatbots; Last update Oct, 2024 |
⚡⚡⚡ Fastest Response 🎯 Simple tasks 💬 Conversation 🖼️ Image Understanding Cost effective via API Limitations: Context (64K tokens), Short responses |
Tips: Use via API , for cost-effective yet intelligent workflows. Otherwise do not bother. Provide concise prompts, examples, use for fast iterations; |
Summarize this product review and highlight key points 📝 Quick text analysis 🎯 Focused responses ⚡ Rapid insights |
| Model & Reasoning Use for Colab coding or don't bother at all | Core Features & Limitations | Best Use Cases & Tips | Example Prompt |
|---|---|---|---|
|
Gemini Pro ++-- Poor performance OK for NLP tasks |
🔗 Might have Multimodal capabilities 💻 Code generation 📝 Text analysis 🖼️ Image understanding Limitations: 32K token context API & Google AI Studio access (Available in supported regions) |
General, modest coding tasks Content creation Image understanding Daily productivity and lifestyle Tips: ✓ Provide clear instructions ✓ Use for balanced tasks ✓ Use for long, less-structured content Will structure and perform NLP tasks economically |
Analyze this image and suggest improvements for the UI design 🖼️ Image understanding 💡 Design insights 📝 Clear explanations Prompting Techniques: Attach or describe the image thoroughly Ask for short, actionable advice Request both text and code solutions |
|
Gemini Ultra + - Reasoning Poor performance, glitchy, and forgetful at times Google's most capable model |
🎯 Some accuracy 🔬 Complex reasoning attempts 📊 Analytics 🖼️ Promised multimodal understanding (not exactly true though) Limitations: Limited availability Enterprise focus (Access through Google Cloud) |
Really good for coding in Colab Tips: ✓ Provide comprehensive context ✓ Repeat prompts when needed ✓ Utilize advanced capabilities of other models |
Analyze this research paper and code implementation, suggesting optimization strategies 📚 Research analysis 💻 Code optimization 🔍 Deep insights Prompting Techniques: Provide full source or summary Ask for step-by-step improvements Request performance metrics |
|
Gemini Classic + - Basic NLP tasks Limited performance; may struggle with retaining instructions Not recommended for complex or demanding tasks |
⚡⚡ Fast response 🗣️ Voice enabled 🎨 Basic DALL‑E integration Limitations: 4,096 tokens |
Routine queries Simple content creation Light tasks Tips: ✓ Use for rapid Q&A ✓ Provide brief context ✓ Ideal for straightforward tasks |
Compare the performance of two simple apps in real time Note: May require repeated instructions due to limited memory ⏱️ Quick comparison 📊 Basic data summary Prompting Techniques: Ask for a brief overview Provide concise context Request a short summary |
