Choosing a Platform
This work is licensed under a Creative Commons Attribution 4.0 International License.
Below are tables that rank the most popular GPTs by popularity from most to least, within each activity type for Chat, Research, Code, and Image/Videos.
Best for Chat¶
| Platform | Strength | Weakness | Cost | Interface | Docs |
|---|---|---|---|---|---|
| Claude 4.0 (Anthropic) | - Fast, coherent dialogue - Good at creative text generation |
- Sometimes less accurate with fact-based queries - Limited third-party integrations |
Free, $15–$75+/mo (various tiers) | Claude Chat | Anthropic Docs |
| Gemini 2.5 (Google) | - Multimodal (images + text) - Strong search integration (Google) |
- Some features only in Beta - Pricing unclear in early versions |
Free, Pro Tiers based on usage | Gemini | Gemini Docs |
| ChatGPT 4.5, o3, o4 (OpenAI) | - Strong reasoning - Integrates with other apps (plugins) - Multi-turn conversation |
- Limited file uploads (depending on plan/plugins) - Occasionally behind newest research releases |
Free, $20/mo (Plus), $42+/mo (Enterprise) | ChatGPT | OpenAI Docs |
| DeepSeek (Open Source) | - Focuses on chat + search integration - Local & open source solutions |
- Smaller dev community - May require self-hosting |
Free (Open Source) | DeepSeek Chat | DeepSeek Docs |
| Microsoft 365 Copilot | - Built on ChatGPT plus Microsoft Graph - Deep MS Office/Outlook integration |
- Typically behind latest ChatGPT features - Enterprise pricing can be high |
$30/mo (add-on for Microsoft 365 E3/E5) | Copilot | Copilot Docs |
| Grok (xAI) | - Multimodal: text, images, basic computer vision - Emerging technology |
- Currently in limited release - Ecosystem still growing |
Free, future subscription TBA | Grok | xAI Docs |
| HuggingFace Chat | - Large variety of community-contributed models - Multimodal options |
- Quality varies by model - Some advanced features behind paid tiers or require GPU credits |
Free, Pay-as-you-go GPU | HF Chat | HF Docs |
| Jasper | - Focus on marketing copy, blog posts - Easy content workflows |
- Less technical depth - Pricier than some for similar text output |
Starting at $39/mo | Jasper Chat | Jasper Docs |
| Perplexity | - Combined chat + search - Large reference library |
- Subscription required for advanced features - Not as flexible as GPT-4 plugins |
Free, Subscription tier $20+/mo | Perplexity Chat | Perplexity Docs |
| NotebookLM (Google) | - RAG (Retrieval-Augmented Generation) - Integrates with Google Drive |
- Still in beta - Occasional slow responses |
Free, future subscription TBA | NotebookLM | NotebookLM Docs |
| Vicuna 13B | - Open-source model fine-tuned from LLaMA - Strong for casual chat |
- Limited knowledge, smaller than GPT-4 - Quality can depend on hosting |
Free (self-host) or free demos | Vicuna Demo | Vicuna GitHub |
| Pi (Inflection AI) | - Empathetic style, personal conversation - Smooth chat experience |
- Less robust for complex tasks - No coding support |
Free (personal use) | Pi | N/A |
| Poe (Quora) | - Aggregates multiple models (GPT-4, Claude, etc.) - Single subscription for multiple AI bots |
- Limited advanced features beyond aggregation - Usage caps on free tier |
$19.99/mo | Poe | Poe Docs |
| Mistral AI | - Open-source European LLMs - Multilingual focus, smaller model sizes |
- Newer offering - Some advanced features still in dev |
Free (open source) or enterprise | Mistral | Mistral Docs |
| Latimer | - Focus on diverse voices - Inclusive perspective |
- Smaller user base - Specialized training set |
Free (basic) or subscription tiers | Latimer | N/A |
| Meta AI (Llama) | - Open source for research/commercial - Large dev community |
- May require significant GPU resources - Some versions less robust vs. GPT-4 |
Free (self-host) or enterprise | Llama | Meta GitHub |
| Apple Intelligence | - Deep Siri/iOS integration - Privacy-oriented design |
- Limited on non-Apple devices - Closed ecosystem |
Included on Apple devices | Apple Intelligence | N/A |
| Amazon Titan | - Part of AWS ecosystem - Integrates with Amazon Bedrock |
- Primarily enterprise usage - Not widely accessible for individuals |
Pay-per-use on Bedrock | Titan | AWS Docs |
| Amazon Bedrock | - Unified API for multiple foundation models - Enterprise security & scaling |
- Must have AWS account - Cost can add up for high usage |
Pay-per-use (AWS) | Bedrock | Bedrock Docs |
| Azure OpenAI Service | - Secure enterprise environment for GPT - Tight Azure cloud integration |
- Requires Azure subscription - Usage cost can be high |
Pay-per-use (Azure) | Azure OpenAI | Azure Docs |
| Merlin AI | - Multifunction: text, image, code - Summaries, chat, quiz creation |
- Some features behind paywall - Web-based aggregator style |
Free plan, paid tiers | Merlin | N/A |
| Starling-7B | - Uses AI feedback (RLHF) - Open-source from UC Berkeley group |
- Smaller model - Less context capacity vs. GPT-4 |
Free (open source) | Starling | Starling GitHub |
| Amplify GenAI | - Open source platform (Vanderbilt) - Connect various model providers |
- Requires AWS + model usage fees - Self-deployment for institutions |
AWS usage + model costs | Amplify GenAI | N/A |
Best for Research¶
| Platform | Strength | Weakness | Cost | Interface | Docs |
|---|---|---|---|---|---|
| Perplexity | - Large archive, direct references - Web search integration |
- Must subscribe for advanced features - Not as “creative” as GPT-like models |
Free, Subscription: $20+/mo | Perplexity Chat | Perplexity Docs |
| Gemini 2.5 Pro (Deep Research) | - In-depth chain-of-thought - Multimodal research capabilities |
- Early product, can be slow - Pricing not fully public |
Free, Pro Tiers based on usage | Gemini | Gemini Docs |
| ChatGPT o1 Pro (OpenAI) | - Strong general-purpose reasoning - Code, language, knowledge tasks |
- Subscription needed | $20/mo (Plus) | ChatGPT | OpenAI Docs |
| Claude 4 Opus (Anthropic) | - Strong reasoning and analysis - Large context window for documents |
- API access can be expensive - Newer than some competitors |
Free, Usage-based | Claude | Anthropic Docs |
| ScholarAI | - Large academic archive integration - Focus on literature reviews |
- Limited reasoning beyond domain texts - Requires GPT-4 or Plus-based backend |
$20–$40/mo (addon to GPT-4) | (No dedicated UI link; typically custom UI) | (Depends on GPT-4 / ScholarAI custom docs) |
| Scholar GPT | - Similar to ScholarAI; academic focus - Pulls from major journal databases |
- Must have ChatGPT Plus - Reliability depends on subscription backend |
$20/mo (Plus required) | (No dedicated UI link; typically custom) | (Leverages GPT-4 docs) |
| Semantic Scholar | - Comprehensive academic metadata - Free to use |
- Not truly an LLM-based chat - Lacks advanced generative capabilities |
Free | Semantic Scholar | Semantic Scholar API |
| Elicit | - AI-assisted literature reviews - Streamlines academic search |
- Limited coverage for some fields - No real-time chat |
Free, Subscription: $10/mo, $42/mo | Elicit | (No official API docs publicly listed) |
| Consensus | - Aggregates academic findings - Summarizes papers via AI |
- May have limited domain depth - Emerging platform |
Free, Subscription: $8-9/mo | Consensus | (No official API docs publicly listed) |
| Scite | - Large academic archive integration - Focus on literature reviews |
- Limited coverage for some fields - No real-time chat |
Free, Subscription: $20/mo | Scite | Scite AI Docs |
| Ai2 OpenScholar | - Answers scientific queries from ~45M open-access papers - Citation-backed responses |
- Focused on open-access content - New & still evolving |
Free | OpenScholar | GitHub |
| Polymathic AI | - Aims at scientific tasks and numerical datasets - Potential for domain-specific fine-tuning |
- Still in development - Limited availability |
Free (open source, in dev) | Polymathic AI | N/A |
| SearchGPT (OpenAI) | - Prototype real-time internet search - Prominent citation linking |
- Waitlist only - Unclear final pricing |
TBD | SearchGPT | N/A |
| You.com | - Search engine + multi-model chat - Customizable “AI Agents” |
- Requires paid subscription for advanced features - Not as robust as GPT-4 in some tasks |
Free or paid plans | YouChat | You.com Docs |
| OpenResearcher | - Retrieves from internet + arXiv - Citation-based conversation for science topics |
- Focused on specific STEM queries - Early-stage tool |
Free, open source | OpenResearcher | GitHub |
Best for Code¶
| Platform | Strength | Weakness | Cost | Interface | Docs |
|---|---|---|---|---|---|
| Claude Code (Anthropic) | - Strong code generation and explanation - Good for complex logic and algorithms |
- Newer than some competitors - API access may have rate limits |
Free, Usage-based | Claude | Anthropic Docs |
| Gemini 2.5 (Google) | - Focus on code + text synergy - Potential for deep chain-of-thought |
- Still early, can be slow with large code bases - Pricing not fully disclosed |
Free, Pro Tiers based on usage | Gemini | Gemini Docs |
| GitHub Copilot | - Seamless code autocompletion - Integrates with VS Code, JetBrains, etc. |
- Limited context window - Some framework-specific gaps |
$10/mo (Individual), $19/mo (Business) | GitHub Copilot | Copilot Docs |
| ChatGPT Code Interpreter | - Interactive code execution in a sandbox - Good for data analysis, transformations |
- Must have GPT-4 subscription - No direct local environment access |
$20/mo (Plus) | ChatGPT | OpenAI Docs |
| Continue.dev | - Open-source local code interpreter - Extensible with plugins |
- Requires technical setup - Smaller dev community |
Free (Open Source) | Continue.dev | Continue.dev Docs |
| Codeium | - Free AI-powered autocomplete - Supports multiple IDEs |
- Less advanced than Copilot on edge cases - Limited large-language reasoning |
Free | Codeium | Codeium Docs |
| Phind | - Code search + AI chat combined | - Smaller community - Some advanced features behind paid tiers |
Free, Paid tiers TBA | Phind | Phind Docs |
| Replit AI | - Cloud-based dev + AI code generator - Multiple languages, strong Python/JS support |
- Usage-based pricing for higher-end features - Emphasis on web IDE only |
Free tier, Pro from $7+/mo | Replit AI | Replit Docs |
| StarCoder | - Open-source code LLM from BigCode - Broad language coverage |
- May lag behind GPT-4/Copilot for advanced logic - Typically self-host or huggingface model usage |
Free (open source) | StarCoder | BigCode |
| Code Llama (Meta) | - Specialized for coding tasks - Multiple variants (Python, Instruct) |
- Less robust than GPT-4 for very complex code - Requires setup for self-host |
Free (open source) | Code Llama | Meta Llama |
Best for Image/Video¶
| Platform | Strength | Weakness | Cost | Interface | Docs |
|---|---|---|---|---|---|
| Veo 3 | - High-quality video generation - Detailed scene and motion control |
- Limited access - High computational cost |
TBD | Veo | N/A |
| Midjourney v6 | - Generally highest-quality AI images - Rich feature set via Discord commands |
- Requires Discord usage - More expensive for casual users |
$10–$60/mo (various tiers) | Midjourney | Midjourney Docs |
| DALL·E 3 (OpenAI) | - Conveniently integrated in ChatGPT - Good for quick generation, concept art |
- Not always photorealistic - Requires ChatGPT Plus or API usage |
$20/mo (ChatGPT Plus) or pay-per-use API | DALL·E 3 | DALL·E Docs |
| Stable Diffusion 3 | - Fully open source - Highly customizable (fine-tuning) |
- Requires GPU resources - Results vary without parameter tuning |
Free (Open Source), or API services | Stability AI | SD3 Paper |
| Adobe Firefly | - Easy to use for designers - Integration with Adobe apps (Photoshop, Illustrator) |
- Requires Adobe subscription - Not as flexible for coding or specialized prompts |
Included w/ Adobe CC or $4.99–$9.99 add-on | Firefly | Firefly Docs |
| Sora (OpenAI) | - Computer Vision (video) generation - Experimental “world simulator” approach |
- Limited early access - High compute demands |
Limited Access, likely enterprise only | Sora | Sora Report |
| Runway ML | - Advanced generative video tools - Browser-based video editor and image gen |
- Higher-res outputs require paid plans - Can be slow on large projects |
Free tier, paid from $15+/mo | Runway | Runway Docs |
| Imagine with Meta | - Text-to-image via Emu model - Free and evolving toolset |
- Factual/historical inaccuracies reported - Still experimental |
Free | Imagine | N/A |
| Craiyon | - Simple web UI - Free usage for low-res images |
- Quality less advanced than Midjourney or DALL·E - Ads on free tier |
Free or subscription | Craiyon | N/A |
About the Table
- ChatGPT is still the dominant chat platform as of late 2024, having recently released model o1 and announcing model o3
- LLaMA 3: Meta's LLaMA 3 is a significant open-source model that is highly competitive and driving innovation.
- Gemini: Google's Gemini (formerly Bard) is rapidly evolving and positioned as a strong competitor to ChatGPT, particularly in multimodal capabilities.
- Copilot: Microsoft's Copilot (integrating OpenAI's technology) has become ubiquitous, especially for productivity and search.
- Claude 3: Anthropic's Claude 3 is gaining popularity due to its strong performance on reasoning tasks and commitment to AI safety.
- Sora: Although it is not yet fully accessible, OpenAI's text-to-video model has the potential to be a major innovation.
- Grok: is powered by what might be the largest super computer in the world, xAI's Collossus
- Image Generation: DALL-E 3, Stable Diffusion 3, and Midjourney v6 are leading the way in image generation, each with its strengths (photorealism, open-source nature, artistic style).
- GitHub Copilot: Remains a dominant force in code generation and is increasingly integrated into developers' workflows.
Tip: If you’re evaluating GPT or generative AI options, consider security (e.g., on-prem vs. cloud), fine-tuning capabilities, and API limits in addition to monthly cost and raw model quality.