Crypto news worth bringing into your next answer

A company announcement can appear across specialist media, financial publications, and public posts. Reading those accounts together helps you see what changed and which details need attention.

Perception organizes tracked sources into searchable mentions. Ask a question, narrow the time window, and keep the source links in your answer.

Our source list is curated by hand. Every outlet is reviewed for editorial quality and relevance to digital assets before it's added. Here's what we cover:

• Crypto-native media. CoinDesk, The Block, Bitcoin Magazine, Blockworks, Decrypt, and CoinTelegraph, among 55 crypto-native outlets. These are the primary sources that break crypto-specific stories.
• Mainstream financial press. Bloomberg, Reuters, Financial Times, Wall Street Journal, CNBC, Forbes, Fortune. Crypto mentions from outlets that move institutional capital.
• Social media. X (Twitter), Reddit, and LinkedIn. Filtered for signal. We track executives, analysts, and key opinion leaders.
• Podcasts and YouTube. Transcribed and searchable. Conference keynotes, panel discussions, and media interviews from channels like Bankless, What Bitcoin Did, and Unchained.
• Regulatory and government. SEC filings, congressional hearing transcripts, enforcement actions, central bank publications, and policy documents.
• Earnings transcripts. Quarterly calls from publicly-traded crypto companies. Strategy (formerly MicroStrategy), Coinbase, Marathon, and 60+ others.
• GitHub repositories. Release notes and significant PRs from 180+ tracked repositories, including Bitcoin Core, Lightning, Ethereum, and major DeFi protocols.

Aggregating news is table stakes. The hard part is making sense of it. Every piece of content that enters Perception goes through an AI enrichment pipeline that adds three layers of structure:

Sentiment scoring. Each article gets a sentiment score from -1.0 to +1.0. The scoring is contextual: it understands that "Bitcoin drops to $60K" is bearish and "Senator drops Bitcoin bill" is neutral. You can filter, sort, and track sentiment over time for any entity or topic.

Entity tagging. NLP entity recognition identifies which companies, people, and protocols are mentioned in each article. This means you can search for "Coinbase" and find mentions even when the article doesn't use that exact word: it might reference "Brian Armstrong's exchange" or "the COIN stock." Over 110 entities are tracked with alias matching.

Narrative extraction. AI trend extraction identifies the broader narratives forming across hundreds of articles. Instead of reading 50 articles about ETF flows, you see "Institutional Bitcoin accumulation accelerating" as a named, tracked trend with signal strength and momentum scores.