{"uid":"cap_MEG8ME31wXI_wbS0aPdO1","slug":"x402-gateway-production-up-railway-app-b3727b68","name":"Perplexity Search-Augmented LLM via x402 Gateway","description":"Search-augmented LLM — answers grounded in real-time web search with citations","url":"https://x402-gateway-production.up.railway.app/api/llm/perplexity","method":"POST","headers":{},"bodySchema":null,"responseSchema":null,"example":{"request":{"messages":[{"role":"user","content":"What are the latest developments in artificial intelligence in 2024?"}]},"response":{"model":"perplexity/sonar-pro","usage":{"cost":0.0129,"is_byok":false,"cost_details":{"upstream_inference_cost":0.0129,"upstream_inference_prompt_cost":0.000042,"upstream_inference_completions_cost":0.012858},"total_tokens":471,"prompt_tokens":14,"completion_tokens":457,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":0,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"content":"In **2024**, the biggest AI developments were the rapid rise of **multimodal and reasoning-capable models**, major advances in **AI regulation**, and a sharp increase in **agentic** and **on-device** AI systems.[2][5][6] The year also saw large new investments in AI infrastructure, with NVIDIA’s **Blackwell** architecture and other chip advances aimed at supporting larger models and faster inference.[1][5]\n\nKey developments included:\n\n- **OpenAI’s GPT-4o** and later **o1** raised the bar for multimodal interaction and reasoning, while competitors such as **Anthropic Claude**, **Google Gemini**, and **Meta Llama** narrowed the gap.[2]\n- **Reasoning and agentic AI** became a major theme, with models increasingly designed to solve tasks step by step and tools like Anthropic’s **Computer Use** public beta showing how models could operate software more directly.[4][7]\n- **Video generation** and other creative media tools advanced significantly, including releases and updates around **Sora**, **Veo**, and image-generation models such as **Imagen 3**.[6][8]\n- **Search and productivity AI** moved into mainstream products, including **ChatGPT Search**, Google’s AI-enhanced search features, and Microsoft’s expanding Copilot ecosystem.[5]\n- **Smaller, local models** grew more important, with companies pushing models that can run on phones and other devices, reducing latency and dependence on cloud compute.[2][4]\n- **AI regulation** became concrete in Europe, where the **EU AI Act** was adopted in March and entered into force on August 1, 2024.[1][5]\n- **AI infrastructure investment** surged, with U.S. private AI investment reaching **$109.1 billion** in 2024 and generative AI drawing **$33.9 billion** globally in private investment.[13]\n\nIf you want, I can also turn this into a **timeline of the top 10 AI moments in 2024** or a **by-company breakdown** of OpenAI, Google, Anthropic, Meta, and NVIDIA."}},"exampleRequest":{"messages":[{"role":"user","content":"What are the latest developments in artificial intelligence in 2024?"}]},"tags":["x402"],"displayCostAmount":"0.060000","displayCostAsset":"USDC","priceDynamic":false,"priceHint":null,"priceStatus":"priced","priceSource":"settled","requiresHandshake":false,"reviewCount":0,"rating":{"score":"0.00","successRate":"0.55","reviews":0,"stars":null,"state":"rated"},"availabilityStatus":"unknown","priceObserved":null,"sessionDeposit":null,"pricing":{"kind":"static","summary":"$0.060000/call","primary":{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.06","per":"call","confidence":"exact"},"accepted":[{"kind":"static","protocol":"x402","network":"base","amountUsd":"0.06","per":"call","confidence":"exact"}]},"paymentMethods":[{"uid":"pm_NMSP5Cj0gkBTLWU5iiK4X","protocol":"x402","methodType":"crypto","chain":"base","mode":"charge","costAmount":"0.06","costPer":"request","priority":0,"asset":null,"unit":"request","depositMicros":null,"planRef":null}],"brandName":null,"brandSlug":null,"brandBaseUrl":null,"brandDocsUrl":null,"whatItDoes":"Answers questions using an LLM grounded in real-time web search, returning responses with citations to live sources","exampleAgentPrompt":"What are the latest developments in the EU AI Act as of this week? I need an answer with citations to real sources, not just your training data.","exampleUseCases":null,"resultDescription":"A natural language answer to the query, grounded in real-time web search results, along with citations including source titles and URLs that back up the response. The answer reflects current information retrieved at query time rather than static training data.","failureModes":["Query times out if web search takes too long — returns error or partial response","No relevant web results found for highly obscure or very niche queries","Payment not included or invalid — returns 402 Payment Required","Perplexity upstream outage — returns 503 or gateway error","Rate limiting if too many requests sent in quick succession"],"whenToPreferThis":"Choose this endpoint when the user's question requires up-to-date, real-world information that a static LLM would not reliably know — such as recent news, current prices, live events, or recent research. It is superior to plain LLM endpoints when source citations are important for trust or verification. Prefer this over raw search APIs when you need a synthesized, readable answer rather than a list of raw URLs.","instructions":null,"reviewSummary":null,"reviewSummaryHighlights":null,"reviewSummaryConcerns":null,"reviewSummaryGeneratedAt":null,"activationCount":11,"lastUsedAt":"2026-05-01T15:00:40.806Z","lastSuccessfullyRanAt":"2026-05-01T15:00:40.806Z","lastHealthCheckAt":"2026-09-15T06:51:02.619Z","isFirstParty":false}