Linkup

Today we announce our $10M seed led by @GradientVC to scale web search for AIs and the launch of Linkup /fast, world’s most accurate sub-second web search API. To celebrate, we’re gifting $100 in query credits for anyone who wants to try, let us know below.

Introducing Linkup Skills. Give your agent built-in prompting best practices – so it generates sharper, more effective queries automatically. It works seamlessly with Claude Code, Cursor, Windsurf, and more. Install in one command: npx skills add LinkupPlatform/skills Link to docs 👉docs.linkup.so/pages/document…

Adding web search to your AI stack shouldn’t mean rebuilding everything 🔨 Linkup’s API + SDK plug right into the tools you’re already using. For example, a lot of teams use @LiteLLM as a proxy to swap keys, handle rate limits, and track usage across their stack. Linkup fits right in – so you can orchestrate everything from one gateway. More on our integrations 👉 docs.linkup.so/pages/integrat…

Linkup is hosting an intimate, technical evening with the Claude Code team next week. If you love building with Claude, come join us for presentations, demos, and good conversations. 📅 Thursday, March, 19 📍 Linkup Offices

AI agents are only as good as the information they can access. 🇪🇺 LightOn, an AI Search & Reason company, is partnering with @Linkup_platform to provide structured, real-time web search to Paradigm within a secure, fully European technology stack. Private data and the open web, securely unified in a single pipeline. Together, the two technologies enable organizations to build AI systems that are better informed, more reliable, and designed for demanding environments, combining search and reasoning to deliver accurate and actionable outputs. 🗞️ Read the press release: lighton.ai/lighton-blogs/…

@Linkup_platform is excited to be part of the v2 of the AI Startup Pack by @Fin_ai. We're joining leading AI-first companies @intercom , @elevenlabs, @attio, and many more to provide AI startups with credits to scale their products. If you’re building an AI-powered startup, check it out: fin.ai/startup-pack

@bymarcoperez @GradientVC Glad you like it. Check your account balance, we have a surprise for you :)

@Linkup_platform @GradientVC I’ve been playing around with your search api the last couple of days. You guys are building something special, wasn’t expecting this great of performance 🤯 Congrats!

Today we announce our $10M seed led by @GradientVC to scale web search for AIs and the launch of Linkup /fast, world’s most accurate sub-second web search API. To celebrate, we’re gifting $100 in query credits for anyone who wants to try, let us know below.

The results? Beyond better token efficiency, they achieved: 3x fewer hallucinated signals 50% lower retrieval latency Every client alert backed by verifiable citations Read the full story here 👉 linkup.so/blog/omniscien…

They implemented Linkup and leveraged two endpoints: i) Deep mode for cited, grounded answers – no leaving the workflow to verify ii) /fetch for clean markdowns when agents need to investigate a specific source Claim verification now runs fully inside the reasoning loop.

Omniscient monitors corporate reputation in real time. One hallucinated signal sent to a client = a liability. Here's how they achieved a 3x reduction in hallucinated signals and cut data retrieval latency by 50% using Linkup’s web search API 🔎

Outcomes for @BareboneTeam : +77% annual returns in 2025 2× faster research workflows Hallucinations mitigated at the retrieval layer Reliable data retrieval is becoming a core component of financial AI systems. Read the full story here: linkup.so/blog/user-stor…

Linkup runs at three layers of Barebone's stack: (i) Backend validation of external data sources (ii) Real-time agent tool calls when live info isn’t in paid data feeds (iii) Personalized morning briefs based on each investor portfolios + watchlists

Here's how one AI investment app reported +77% returns in 2025. That’s 4.7× the S&P 500 and 2x the performance of leading hedge funds. A key part of their system: agents that detect emerging market narratives before they reach traditional media. Here’s how their stack works👇

Financial AI has no tolerance for bad data. A hallucinated number or misquoted filing quickly turns into a bad trade. Their previous tools returned static data points data points, struggled with parallel queries, and broke under scale. Linkup provided their agents with real-time web search 👉 high-precision, scalable, reliable retrieval.

On Monday we hosted GTM Engineering After Dark with @clay, @joinhexa, and @fullenrich 🌙 30+ GTM leaders from top EU startups & scale-ups joined us to build enrichment waterfalls and custom signals. Our COO also walked through one of the most common Clay integrations powered by real-time web data. A few highlights from the night: 1) APIs are everywhere – GTM Engineers are leaning more into their engineering role and coding heavily (especially with tools like Claude Code) 2) Clay is uniquely powerful, and becomes even more valuable when combined with external APIs 3) You can now build outbound sequences that go far beyond “I saw you liked a post on LinkedIn…” by combining multiple signals and real-time data Big thanks to the @clay team for flying in from the US and to @joinhexa for hosting 🙏 Learn more about our Clay integration 👇 docs.linkup.so/pages/integrat…

Full methodology, limitations, and open-source evaluation code are linked below. We're sharing this so teams can run the same tests on their own queries. 👉 linkup.so/blog/evaluatin…

Results varied significantly across all three dimensions: → Up to 3× difference in unique sources returned per query → Up to 4× difference in missing-entity rate → Hallucination rate showed the widest variance

We benchmarked 4 AI search APIs on queries drawn from real user traffic. The goal was to understand why our users choose Linkup over competitors and to measure what actually matters in a professional search workflow. Blind judging with LLMs. Open-source methodology. Here's what the data showed 👇