The AI landscape shifts fast. We’ve scraped Hacker News, Reddit, and ProductHunt to surface the most significant launches and updates from the past week. Here’s what actually matters.
—
Quick Verdict
> Claude leads the pack this week with the highest total Hacker News engagement (20,424 points, 11,445 comments), signaling major user interest in its latest capabilities. DeepSeek is the dark horse for cost-conscious developers at $0.27/M tokens API pricing. ChatGPT remains the most versatile generalist, but Gemini is closing the gap with aggressive updates.
—
The New LLMs: Claude, ChatGPT, Gemini, DeepSeek
Claude (Anthropic)
What it does: Claude continues to dominate discussion with 20,424 HN points and 11,445 comments. The latest updates focus on improved reasoning, longer context windows, and better instruction following.
Why it matters: Claude’s safety-first approach resonates with enterprise users who need reliable outputs without hallucinations. The community response suggests Anthropic is shipping features that directly address developer pain points.
Who it’s for: Teams building customer-facing chatbots, legal document analysis, or any application where accuracy and safety are non-negotiable.
Pricing: Free tier available. Pro at $20/month. Team at $30/user/month.
ChatGPT (OpenAI)
What it does: With 14,379 HN points and 8,038 comments, ChatGPT remains the most widely used AI assistant. Recent updates include improved multimodal capabilities and faster response times.
Why it matters: OpenAI’s aggressive feature rollout keeps ChatGPT as the default choice for most users. The ecosystem of plugins and GPTs continues to expand.
Who it’s for: General users, content creators, and developers who want the broadest set of capabilities in one tool.
Pricing: Free tier available. Plus at $20/month. Team at $25/user/month.
Gemini (Google)
What it does: Gemini scored 15,813 HN points with 7,287 comments. Google’s latest updates integrate Gemini deeper with Workspace and Android, making it the most embedded AI assistant in daily workflows.
Why it matters: Google’s advantage is distribution. Gemini is already in Gmail, Docs, and Android. For users already in Google’s ecosystem, the switching cost is zero.
Who it’s for: Google Workspace users, Android users, and anyone who wants AI assistance without leaving their existing tools.
Pricing: Free tier available. Advanced at $19.99/month. API pricing varies.
DeepSeek
What it does: DeepSeek continues to impress with 12,067 HN points and 6,066 comments. The Chinese AI lab’s latest models offer competitive performance at a fraction of the cost.
Why it matters: At $0.27 per million tokens for API access, DeepSeek is the cheapest option for high-volume inference. The open-weight models allow self-hosting.
Who it’s for: Budget-conscious developers, researchers, and teams building cost-sensitive applications.
Pricing: Free tier available. API at $0.27/M tokens. No team pricing — API only.
—
Developer Tools: GitHub Copilot, Cursor AI, Codeium, Tabnine
GitHub Copilot
What it does: 10,420 HN points and 6,456 comments reflect Copilot’s dominance in AI-assisted coding. The latest updates include better multi-file editing and improved context awareness.
Why it matters: Copilot is now integrated into VS Code, JetBrains, and Neovim. For most developers, it’s the default AI pair programmer.
Who it’s for: Any developer writing code daily. The free tier makes it accessible for students and hobbyists.
Pricing: Free tier available. Individual at $10/month. Business at $19/user/month.
Cursor AI
What it does: 1,043 HN points with 585 comments. Cursor is an AI-first code editor built on VS Code. It offers features like AI-powered code review and natural language commands for refactoring.
Why it matters: Cursor is the fastest-growing alternative to Copilot for developers who want deeper AI integration in their editor.
Who it’s for: Developers who want AI built into the editor itself, not just as a plugin.
Pricing: Check website for current pricing.
Codeium
What it does: 273 HN points, 144 comments. Codeium offers free AI code completion for individual developers. It supports 70+ languages and 40+ IDEs.
Why it matters: The free tier is genuinely useful. For solo developers who can’t justify $10/month for Copilot, Codeium is a solid alternative.
Who it’s for: Individual developers, students, and teams on a tight budget.
Pricing: Check website for current pricing.
Tabnine
What it does: 660 HN points, 199 comments. Tabnine focuses on privacy-first code completion. It can run entirely on-premises, making it the choice for regulated industries.
Why it matters: For companies that cannot send code to cloud APIs, Tabnine’s local inference is the only viable option.
Who it’s for: Enterprise teams in finance, healthcare, and government with strict data residency requirements.
Pricing: Check website for current pricing.
—
Productivity AI: Perplexity, Notion AI
Perplexity
What it does: 4,766 HN points, 2,682 comments. Perplexity is an AI-powered search engine that cites sources. The latest updates improve answer accuracy and add deeper research capabilities.
Why it matters: Traditional search is broken for complex queries. Perplexity’s citation-based answers save hours of manual research.
Who it’s for: Researchers, journalists, and anyone who needs accurate, sourced information quickly.
Pricing: Free tier available. Pro at $20/month. Team at $40/month.
Notion AI
What it does: 928 HN points, 459 comments. Notion AI adds writing assistance, summarization, and knowledge management to Notion’s existing workspace platform.
Why it matters: For teams already using Notion, the AI features reduce friction in documentation and note-taking.
Who it’s for: Teams using Notion for project management and documentation.
Pricing: Check website for current pricing.
—
Comparison Table
| Tool | Category | HN Points | HN Comments | Starting Price | Best For |
|——|———-|———–|————-|—————-|———-|
| Claude | LLM | 20,424 | 11,445 | $20/month | Safety-critical applications |
| DeepSeek | LLM | 12,067 | 6,066 | $0.27/M tokens | Cost-sensitive developers |
| ChatGPT | LLM | 14,379 | 8,038 | $20/month | General AI assistance |
| Gemini | LLM | 15,813 | 7,287 | $19.99/month | Google ecosystem users |
| Perplexity | Search | 4,766 | 2,682 | $20/month | Research and fact-checking |
| GitHub Copilot | Developer | 10,420 | 6,456 | $10/month | Daily coding |
| Cursor AI | Developer | 1,043 | 585 | TBD | AI-first coding |
| Codeium | Developer | 273 | 144 | TBD | Free code completion |
| Tabnine | Developer | 660 | 199 | TBD | Privacy-focused coding |
| Notion AI | Productivity | 928 | 459 | TBD | Documentation |
—
User Sentiment
What users love:
– Claude gets praised for honesty — it admits when it doesn’t know something rather than hallucinating.
– DeepSeek is called “the best value in AI” by developers running high-volume workloads.
– Gemini integration with Google Workspace is a genuine time-saver for heavy Docs users.
– GitHub Copilot is “good enough to be worth $10/month” for most developers.
What users hate:
– ChatGPT still has reliability issues with complex multi-step instructions.
– Perplexity sometimes cites irrelevant or low-quality sources.
– Codeium free tier has noticeable latency compared to paid alternatives.
– Tabnine local models are less capable than cloud-based competitors.
—
Who Is Each Product For?
Claude: Teams building customer-facing AI where hallucinations are unacceptable. Legal, medical, and financial applications.
DeepSeek: Developers running high-volume inference on a budget. Self-hosting enthusiasts.
ChatGPT: General users who want the widest range of capabilities. Content creators, marketers, and casual users.
Gemini: Google Workspace power users. Android users who want AI baked into their OS.
Perplexity: Researchers, journalists, and students who need verified, sourced information.
GitHub Copilot: Professional developers writing code daily. The best value for active coders.
Cursor AI: Developers who want AI deeper in their editor than Copilot provides.
Codeium: Students and hobbyists who need free code completion.
Tabnine: Enterprise teams that cannot send code to external APIs.
Notion AI: Teams already using Notion who want AI-assisted documentation.
—
FAQ
Which AI tool has the best free tier?
ChatGPT’s free tier is the most generous for general use. DeepSeek offers the cheapest API pricing at $0.27/M tokens for developers.
Is Claude better than ChatGPT for coding?
Not yet. GitHub Copilot and ChatGPT still lead for code generation. Claude excels at reasoning and safety, but code-specific tasks favor dedicated tools.
Can I run these models locally?
DeepSeek and Tabnine offer local deployment options. Claude, ChatGPT, and Gemini are cloud-only.
Which tool is best for research?
Perplexity’s citation system makes it the clear winner for verified research. Claude is a close second for complex analytical tasks.
—
[IMAGE PROMPT: photorealistic top-down desk setup featuring a laptop with AI tool interfaces visible on screen, alongside a smartphone and notebook, natural lighting from a window, clean modern desk, no text or logos]
—
SoftRanked is reader-supported. When you buy through links on our site, we may earn an affiliate commission.
Last updated: July 17, 2026