Finacial x data gatherer
Prompt
# AI Benchmark: Real-Time NVIDIA Financial Intelligence You are participating in a competitive benchmark evaluating which AI model can retrieve the **fastest, freshest, most accurate, and most reliable financial intelligence** from X (formerly Twitter) and other financial information sources. All participating AI models receive the same instructions. Your performance will be compared against competing models using independently verified results. ## 1. Objective Retrieve the latest available information concerning **NVIDIA Corporation (NASDAQ: NVDA)** as of the exact moment you begin your research. **Research date:** 17 November 2026 **Timezone:** Europe/Amsterdam (CET, UTC+1) **Information cutoff:** Your actual research start timestamp, recorded to the second. Your primary source is X, but you should independently verify important information using reputable financial data providers, exchanges, official company announcements, and primary reporting. Your goal is to demonstrate the greatest possible information freshness, source accuracy, and retrieval efficiency. ## 2. Research Requirements ### A. Real-Time X Intelligence Search X as comprehensively as your available tools permit for the latest relevant information about: - NVIDIA stock prices, premarket activity, and price movements. - Breaking financial news concerning NVIDIA. - Earnings, revenue forecasts, and financial guidance. - Analyst upgrades, downgrades, and price-target changes. - Institutional activity and significant transactions. - AI semiconductor industry developments that could materially influence NVDA. - Regulatory decisions, export restrictions, and geopolitical developments. Find the **10 most recent materially relevant original X posts**, prioritizing freshness, credibility, and direct market relevance. Do not confuse the newest posts with the most popular posts. ### B. Latest Stock Price Identify the most recent verifiable NVIDIA stock price available at the cutoff. Report: - Exact price in USD. - Absolute and percentage change against the appropriate previous close. - Market session: premarket, regular, or after-hours. - Exact quote timestamp, including timezone. - Data provider and whether the quote is real-time or delayed. - Whether the price is independently confirmed. Distinguish an actual market quote from a price merely mentioned in an X post. ### C. Breaking-News Detection Identify the five most important developments concerning NVIDIA that were available at the cutoff. For each development, determine: - What happened. - When the information first became publicly available. - Who originally reported it. - Whether later reports simply repeated earlier information. - Whether it is verified, disputed, or speculative. - Why it could influence NVIDIA's valuation or market sentiment. Prioritize **early detection of genuinely new information**, not simply summarizing already widely circulated news. ## 3. Mandatory Source Verification For every X finding, provide: 1. Author's name and X handle. 2. Direct URL to the original post. 3. Exact original publication timestamp. 4. Age of the post in seconds or minutes at the research cutoff. 5. Concise description of the information. 6. Original reporting, repost, commentary, or speculation classification. 7. Independent verification status. Never invent URLs, timestamps, prices, posts, or quotations. If a source's publication time cannot be verified, mark it as unknown. Do not estimate it. If direct X access is unavailable, disclose the limitation and identify any alternative retrieval method. **A model that admits missing information should score higher on reliability than a model that fabricates convincing results.** ## 4. Retrieval Speed and Efficiency Record the following timestamps: - Research start. - First relevant result retrieved. - First independently verified result. - Research completion. Calculate: - Time to first relevant result. - Time to first verified result. - Total research duration. - Median and minimum age of the retrieved posts at the cutoff. Distinguish actual measured timing from estimates. Do not claim millisecond or second-level precision unless your tools provide that information. Efficiency matters, but speed must not come at the expense of accuracy. ## 5. Strict Benchmark Conditions - Use only information available at or before your recorded cutoff. - Do not rely on cached knowledge while presenting it as live research. - Do not claim to have searched X directly if you accessed only indexed results. - Do not fabricate access to APIs or financial terminals. - Do not use subsequently published information to improve your answer. - Do not substitute unrelated general NVIDIA news to fill missing results. - Clearly distinguish rumors from verified developments. - Do not claim complete X coverage without evidence. - State which retrieval tools and source types were actually used. - Do not present financial predictions as established facts. - Do not optimize your answer for length. Optimize for information quality. ## 6. Required Output ### A. Research Metadata Provide the precise research start timestamp, cutoff, completion timestamp, total duration, and retrieval methods. ### B. Latest Verified NVDA Price Present the latest price, change, session, original quote timestamp, provider, and verification status. ### C. Ten Latest X Findings Create a table ordered by original publication time, newest first. | # | Time (CET) | Age at cutoff | X account | Finding | Verification | Direct URL | |---|---|---|---|---|---|---| ### D. Five Most Important Market Developments Rank developments by potential market significance, not social-media popularity. Provide supporting sources and clearly identify uncertainties. ### E. Retrieval Performance Report measured speed, information freshness, source coverage, access limitations, and any information that could not be verified. ### F. Self-Assessment Provide a confidence score from 0–100, explaining the evidence and limitations behind it. This score is not the final benchmark score. ## 7. External Benchmark Scoring Your response will be independently evaluated using the following weights: | Category | Weight | |---|---:| | Information freshness | 30% | | Factual accuracy and independent verification | 25% | | Direct X access and valid original post URLs | 20% | | Financial news coverage and relevance | 10% | | Retrieval speed | 10% | | Transparency and limitations | 5% | | **Total** | **100%** | Fabricated financial prices, source URLs, or timestamps will receive substantial penalties regardless of performance in other categories. ## Final Instruction **Begin your research immediately.** Your objective is to identify which AI system is best suited to serve as a real-time financial intelligence agent within a multi-agent trading organization. Do not provide trading recommendations or investment advice. Demonstrate your capabilities through **actual retrieval, verifiable evidence, source freshness, and measured performance**, rather than explanations about what you theoretically could retrieve.