Octen, the fastest search for AI, today launches Octen Pulse, a generative retrieval model that returns search results in 10 milliseconds. At that speed, search stops being a separate step a model pauses for and becomes part of how it reasons. Octen is also introducing a $20-a-month subscription that gives individual developers unlimited API calls.
Pulse is open source and joins Octen Fast and Octen Premier as the third search method in Octen’s lineup. The subscription also includes image search, expanding Octen into multimodal search one year after the company’s founding. Since announcing general availability in July, Octen has grown to 5,000+ active developers.
Most AI search APIs charge per call, and agents search dozens of times per task, so the bill climbs with every step. Developers respond by letting their agents search less, and the answers get worse. Speed adds a second cost. Most APIs built for AI take more than 300 milliseconds per call, so a model finishes a reasoning step, stops, searches and waits. Octen’s subscription removes the cost trade-off, and its search methods remove the wait.
Three Ways to Search
- Octen Pulse is a new generative retrieval model that returns results in 10 milliseconds. Pulse has learned millions of URLs across major websites, so a single inference turns a question directly into a matching result. Each search takes about the same time and cost as generating one token, which lets retrieval run inside a model’s reasoning. Because Pulse is open source and can run on-device, it also brings real-time search to physical AI, including robots, smart glasses and smartphones that can’t wait on a round trip to the cloud.
- Octen Fast is Octen’s core search API. It returns structured results at a median of 62 milliseconds on the SealQA Hard benchmark, more than 4x faster than its nearest rival. In the Artificial Analysis Search API Benchmark, Octen was the fastest provider at 0.21 seconds per search query and 15.1 seconds per task. Fast supports more than 1 million queries per second.
- Octen Premier handles specialized searches, including business search, news search and place search, through the API and the Octen console.
“Every time the internet got faster, a wave of new applications followed, and search for AI works the same way,” said Kuan Zou, founder and CEO of Octen. “How fast a model can retrieve information decides how fast intelligent products can ship, which is why we’re so aggressive about speed. Once a product reaches real scale, cost matters just as much. We’re building search the way utilities deliver water and electricity, so developers never have to ration it as they grow.”
Unlimited Search for $20 a Month
Octen’s subscription brings the pricing model developers already know from LLMs to search. For $20 a month, individual developers get unlimited API calls, so agents can search as often as a task needs.
Octen also performs well on cost in independent testing. Artificial Analysis runs the same model and harness against 26 search API products from 13 providers. In its September results, Octen’s search cost $9.07 per 1,000 tasks, compared with $30 to $66 for other providers.
“For agents that rely on the web, search can sometimes cost more than the model itself. What we like about Octen is that it makes search spend easier to predict, so developers can focus on when their agents need to look something up and spend less time worrying about the bill,” said Sid Sharma, Head of GTM at Inception.
Image Search for Multimodal AI
Octen’s subscription now includes image search. Image search gives image generation models real visual references to ground what they create. It also lets agents build answers that combine text and images. An agent comparing products, for example, can pull product photos through the same service it uses for text, then find and present the information in one step.
“An LLM is like a powerful brain. Its true potential is unlocked when it can access the world’s latest information instantly and explore knowledge at scale,” said Professor Aixin Sun, Associate Dean of the College of Computing and Data Science at Nanyang Technological University. “Intelligence becomes far more powerful when information is abundant, current, and immediately available.”
A Team That Has Built Search at Scale Twice
Zou previously led enterprise search at Baidu and AI search at Alibaba Cloud. At Alibaba, he helped build search infrastructure that handled some of the world’s largest instantaneous traffic spikes, including those during Singles’ Day. At Octen, he built his first search engine for machines. His team comes from Meta, Google, TikTok, Alibaba, Baidu, DeepSeek and Xiaohongshu, companies that run the world’s largest search and recommendation systems. The company is focused on preparing search infrastructure to meet demand for machine-generated queries, making search more reliable, affordable, and accessible.
Developers and AI teams can get an API key and run their first queries at octen.ai.
About Octen
Octen is the search engine built for machines. Its API gives AI models and agents search across text and images, with text results in 10 milliseconds, and ranks results by relevance to the ground truth rather than predicted clicks. Founder and CEO Kuan Zou previously led AI search at Alibaba Cloud and enterprise search at Baidu. Octen is backed by Square Peg and Argor. The company is headquartered in San Francisco.
View source version on businesswire.com: https://www.businesswire.com/news/home/20261113861590/en/
Media gallery
