The Era of AI Execution Has Arrived: 31 Most Promising AI Application Companies Worldwide
TechFlow Selected TechFlow Selected
The Era of AI Execution Has Arrived: 31 Most Promising AI Application Companies Worldwide
AI is no longer just an "assistant," but an "executor."
Navigating Web3 tides with focused insightsBy Yu Ji
Today, Mr. Crow brings you a look at the world's top 31 AI application companies.
In April 2025, Forbes released its seventh annual AI 50 list. Compiled by Forbes in collaboration with Sequoia Capital and Meritech Capital, this ranking highlights the 50 most promising private artificial intelligence companies globally.
Among them, AI application companies stand out as the biggest highlight—31 of the 50 listed firms fall into this category.
Besides well-known names like Perplexity, a new wave of AI application startups has emerged, including Anysphere (creator of the AI coding tool Cursor), which helps engineers write code, and Speak, an AI-powered language tutoring app.
From these shifts among AI application companies, Sequoia Capital identified two key trends:
First, AI is moving from "chatting" to "acting": 2025 marks a turning point for AI applications. While AI was once limited to answering questions or generating content, it now actively completes full workflows. AI is no longer just an "assistant"—it’s becoming an "executor."
Second, enterprise-grade AI tools are taking center stage: Legal tech company Harvey can now automate entire legal processes—from document review to client communication—effectively replacing an entire team of junior lawyers. Similarly, Sierra and Cursor have made the leap from "assisting" to "autonomously completing tasks" in customer service and software development, respectively.
Now, let’s dive into the list and explore which AI companies made the cut.
/ 01 / Perplexity AI: The Search Powerhouse of the AI Era
Perplexity AI is an AI-powered search engine and web search tool focused on developing advanced text generation and comprehension technologies.
The company specializes in building conversational search engines based on large language models (LLMs), aiming to deliver accurate, real-time answers with citations from reliable online sources.
Perplexity AI simplifies research and information retrieval by automatically collecting, analyzing, and summarizing data from current sources, academic papers, and trusted databases. It is especially useful for students, researchers, journalists, and professionals who need fast, reliable answers without manually sifting through multiple sources.
As of December 2024, Perplexity AI had reached a valuation of $9 billion, with annual revenue of approximately $350 million.
/ 02 / Figure AI: Don’t Feel Like Working? Let the Robot Do It!
Figure AI is a robotics company developing AI-driven humanoid robots capable of performing physical tasks in complex real-world environments such as warehouses, factories, retail stores, and even homes.
Figure AI aims to reduce human labor and fill roles that “humans don’t want to do,” addressing the growing labor shortage crisis in the United States.
Its flagship product, Figure 02, is a robot equipped with voice dialogue capabilities, visual reasoning, and highly dexterous robotic hands. It has already been deployed at BMW’s factory in South Carolina, where it performs assembly-line tasks.
By early 2025, Figure AI was raising $1.5 billion in funding, reaching a valuation of $39.5 billion.
/ 03 / Glean: Enterprise Search Powered by AI
Glean focuses on developing AI-powered enterprise search chatbots, bringing AI into core business workflows.
In today’s information-rich era, many companies struggle with data overload. Employees face massive volumes of unstructured information daily, making data organization a major challenge. Departmental silos further create knowledge barriers, hindering efficient access to internal corporate knowledge.
Leveraging integrations with over 100 SaaS applications, Glean aggregates company-wide information to deliver comprehensive results. By connecting to internal data sources via AI and workplace search platforms, it builds enterprise knowledge graphs that help employees find information more effectively.
Glean also enables personalized searches for individual employees, offering targeted interactions. As an AI assistant, it enhances internal information flow and improves communication efficiency across teams.
After closing a $260 million Series E round in September of last year, Glean reached a valuation of $4.6 billion.
/ 04 / Runway: The AI Behind 'Everything Everywhere All at Once'’s Visual Effects
Runway is an AI video and image editing software—and a global leader in AI-generated video. Its professional-grade output surpasses other AI video tools.
Centered around “AI + visual creation,” Runway offers more than 30 AI-powered video and image tools and models. Through continuous model iteration and technological innovation, it delivers film-quality visual effects.
In 2023, *Everything Everywhere All at Once* won seven Oscars, including Best Picture. Despite featuring extensive visual effects, the film’s VFX team consisted of only five people!
This wasn’t achieved through grueling overtime but rather by leveraging Runway to produce stunning, complex effects with minimal manpower—boosting efficiency, cutting costs, and enhancing creative possibilities. A true triple win!
Runway’s pricing is also user-friendly: $12/month for the individual Pro plan, and $28/month for teams.
In July 2024, Runway secured $450 million in Series D funding, achieving a $4 billion valuation.
/ 05 / Harvey: Streamlining Legal Workflows
AI has entered the legal field—and Harvey is leading the charge. This AI-powered legal advisor application leverages OpenAI’s GPT-4 model to understand natural language inputs from lawyers and rapidly generate legal arguments and draft documents.
Complex legal procedures that once took hours can now be completed by AI in seconds, serving as an excellent support tool that dramatically boosts lawyer productivity.
In July 2024, Harvey closed a new funding round, reaching a $1.5 billion valuation.
/ 06 / Abridge: The AI Medical Scribe
Abridge stands out as one of the most representative players in the AI healthcare space. This AI medical note-taking software uses AI models to transcribe conversations between patients and clinicians, organizing them into structured clinical notes.
In the U.S., doctors are required to document every patient consultation. Due to the complexity and volume of work, a dedicated role—medical scribes—emerged. There are about 100,000 scribes in the U.S., accounting for roughly 10% of all physicians—an enormous labor cost.
Abridge is now using powerful AI models and algorithms to replace this workforce. Compared to traditional scribes charging $25/hour, Abridge reduces transcription costs to just $0.12/hour.
Today, Abridge partners with U.S. healthcare systems and is integrated into official clinical workflows, showing immense potential. In February 2025, it raised $250 million, reaching a $2.75 billion valuation.
/ 07 / DeepL: Professional AI Translation That Speaks for Itself
DeepL is an AI-powered language translation software renowned for its high accuracy and rich feature set, setting it apart from competitors.
Powered by specialized AI translation models, DeepL reportedly delivers translations three times more accurate than its closest rivals.
It supports automatic recognition of 33 languages and offers voice translation via microphone. Beyond text, it handles file formats like PDF and PPT.
In 2024, DeepL launched Deepl Write Pro—an AI writing assistant designed to refine and enhance text, improving professionalism and clarity.
Over 100,000 global enterprises use DeepL—including half of the Fortune 500 companies—such as Zendesk, Nikkei News, Coursera, and Deutsche Bahn.
In May 2024, DeepL raised $300 million in Series C funding, reaching a $2 billion valuation.
/ 08 / Notion: Everything in Notion
Notion is an all-in-one productivity suite combining notes, task management, databases, and collaboration. With a modular design, users can freely combine blocks of text, images, tables, code, and videos—like LEGO bricks—to build personalized pages and workflows.
It supports cross-platform use (Linux/Mac/Windows/iOS/Android) and real-time multi-user collaboration. Thanks to seamless syncing and backup features, users enjoy smooth transitions across devices—a powerful productivity tool for both individuals and teams.
By mid-2024, Notion surpassed 100 million registered users worldwide—a milestone reflecting widespread adoption and solidifying its leadership in the productivity software market.
In October 2021, Notion raised $275 million in Series D funding, achieving a $10 billion valuation.
/ 09 / Synthesia: Creating Virtual Humans More Human-Like Than Humans
Synthesia is a leading AI video generation platform that creates lifelike avatars and speech, supporting over 100 types of AI-generated virtual presenters and 120+ languages. It pioneered the ability to create virtual characters without any filming equipment.
Synthesia’s core capabilities include generating realistic AI avatars, natural-sounding voices, and high-quality video content.
Powered by its EXPRESS-1 model, Synthesia produces AI avatars whose voices, intonations, facial expressions, and movements increasingly resemble real humans—blurring the line between virtual and reality.
On January 15, 2025, Synthesia closed an $180 million Series D round led by New Enterprise Associates.
/ 10 / WriterLabs: The Marketing Expert That Knows Your Company Better Than You Do
WriterLabs is a generative AI platform tailored for enterprise marketing. It helps teams boost content creation efficiency, maintain brand consistency, and accelerate workflow automation.
Using its proprietary large language model, Palmyra, WriterLabs automatically generates marketing copy aligned with a company’s brand voice and style, delivering innovative, customized content.
Unlike general-purpose AI writing tools, WriterLabs’ edge lies in its in-house developed Palmyra model—trained on 1 trillion tokens and fine-tuned for specific domains like healthcare and law, ensuring greater industry relevance, professionalism, and versatility.
The company claims Palmyra outperforms competitors like ChatGPT in accuracy and effectiveness, while offering lower API costs—reducing overall expenses.
As of November 2024, WriterLabs raised $190 million in Series C funding, reaching a $1.9 billion valuation, with clients including Uber, Salesforce, and Spotify.
/ 11 / Skild AI: Building the Brain for Robots
Skild AI is a Pittsburgh-based startup focused on developing an AI “robot brain”—Skild Brain—that enables robots to perform diverse tasks in complex, dynamic environments, such as object manipulation, navigation, and motion control.
Unlike traditional robots built for specific tasks, Skild Brain serves as a universal, shared intelligence that can power various robots across different scenarios.
The company claims its model is trained on at least 1,000 times more data points than competing models, enabling superior generalization so robots can operate effectively in unseen environments.
With global labor shortages expected to result in 85.2 million unfilled jobs by 2030—costing $8.5 trillion in lost economic output—Skild AI is positioned to capture significant market opportunity.
By July 2024, Skild AI had completed a $300 million Series A round, achieving a $1.5 billion post-money valuation.
/ 12 / World Labs: Bringing AI Into the 3D Era
World Labs is a spatial intelligence application that uses Large World Models (LWMs) to enable AI to perceive, generate, and interact with 3D environments. With just one image, it can generate a 3D version; with just one sentence, it can bring mental imagery into tangible 3D form.
Its core technology transforms 2D input images into interactive 3D scenes, elevating flat graphics into immersive spatial experiences. By constructing interactive 3D spaces, users gain intuitive understanding of complex real-world objects and situations.
Even more impressively, World Labs doesn't just generate 3D worlds—it enables AI to think and reason within them.
Founded by Dr. Fei-Fei Li, known as the “Godmother of AI” and a professor at Stanford University, she believes perception and action are deeply intertwined. Only when models perceive in 3D can they act in 3D—this is a crucial inflection point in AI evolution.
By September 2024, World Labs had raised $230 million, achieving a valuation exceeding $1 billion. The round was co-led by Andreessen Horowitz, New Enterprise Associates, and Radical Ventures, with participation from Intel Capital, Marc Benioff (personally), and Ronald Conway (personally).
/ 13 / Sierra: Hello, I’m Your Dedicated AI Phone Assistant
Sierra is a customer service software that handles inquiries and voice calls, delivering personalized, brand-consistent conversational experiences. It specializes in building conversational AI platforms for enterprises to improve service efficiency and quality.
Sierra’s technical architecture combines multiple models—including those from OpenAI, Anthropic, and Meta—to ensure accuracy and reliability. It also employs a “supervised agent” mechanism, using real-time monitoring and quality controls to minimize AI hallucinations and maintain trust in customer interactions.
As of October 2024, Sierra had raised $175 million, reaching a $4.5 billion valuation, with clients including WeightWatchers, SiriusXM, and Sonos.
/ 14 / ElevenLabs: Clone Voices in Seconds—Indistinguishable from Reality!
ElevenLabs is a company specializing in AI-powered voice generation and voice cloning, founded in 2022. Its core technology enables rapid creation of natural, fluent speech and precise replication of specific voices.
Its main features include:
Voice Cloning: Users upload 30 seconds to 5 minutes of audio to replicate a person’s voice—including tone, accent, and even emotional nuances.
Voice Synthesis: After cloning, users input text to generate natural-sounding speech.
Bidirectional Conversion: Beyond text-to-speech, ElevenLabs also supports speech-to-text transcription, enabling full bidirectional conversion.
As of January 2025, ElevenLabs completed an $180 million Series C round, reaching a $3.3 billion valuation.
/ 15 / Windfall: Determining the Net Worth of Every Person on Earth
Windfall’s mission is “to determine the net worth of every individual on Earth.” It is an AI-powered intelligence analysis software that processes personal data to provide organizations with deep insights, helping them better understand target and prospective customers to drive more effective marketing and engagement.
However, Windfall primarily serves nonprofit organizations. By providing weekly-updated, reliable consumer financial data, it helps nonprofits identify and attract donors.
Windfall compiles public database information—such as property values, mortgages, occupation details, and past donations—that may indicate a person’s giving capacity. Using data-driven insights, it identifies, segments, and engages audiences to maximize fundraising outcomes.
/ 16 / Anysphere: Making Programming No Longer a Challenge
Anysphere is an innovative company focused on automated coding, founded in 2022 in San Francisco. Its flagship product is Cursor, a critically acclaimed AI-powered code editor.
Cursor is currently one of the leading programming tools in the AI space. It converts natural language instructions into precise code, answers coding-related questions, and detects and fixes bugs instantly.
Cursor offers three AI interaction modes: In Agent mode, AI fully takes over coding—ideal for beginners; Edit mode gives experienced developers control while AI assists; Ask mode focuses on Q&A, letting users consult AI on any coding issue.
In 2025, Cursor raised $625 million, reaching a $9.6 billion valuation, led by Thrive Capital and a16z, with Accel joining as a new investor.
/ 17 / Luminance: One-Click Legal Contract Generation
Luminance is a leading UK-based legal tech company focused since 2016 on AI solutions for contract management, review, and negotiation.
Developed by AI experts from Cambridge University, Luminance’s platform uses a proprietary legal large language model called LegalPre-trainedTransformer (LPT) to automatically generate, review, and analyze legal contracts.
Unlike general-purpose models like GPT, LPT is trained on over 150 million verified legal documents—many of them non-public—ensuring high accuracy and reliability for legal professionals.
Luminance can quickly analyze vast volumes of documents, identifying potential legal risks and clauses, drastically reducing time and effort compared to manual review. It also provides intelligent compliance checks, helping businesses ensure their contracts and operations meet regulatory requirements.
In early 2025, Luminance announced a $75 million Series C round led by Point72 Private Investments—one of the largest financings ever in the UK and European pure-play legal AI sector.
/ 18 / Speak: AI Helps You Sound Like a Native Speaker
Speak, an AI learning startup based in San Francisco, is a newcomer on this year’s list.
Founded in 2016, Speak is an AI-powered spoken language learning app that uses AI—particularly OpenAI’s GPT-4—to help users learn new languages like native speakers.
Language learning typically involves four components: listening, speaking, reading, and writing. Most native speakers acquire language through listening and speaking first. Speak mimics this approach, structuring learning into three stages.
First, users practice listening and speaking. Then, Speak reinforces newly learned vocabulary or phrases through repetition. Finally, AI simulates real-life contexts, allowing users to experience firsthand how language and words are used naturally.
Currently, users can only learn English and Spanish on Speak, with French planned for release soon. The app already has over 10 million users and operates in over 40 countries.
In December 2024, the company raised $78 million in a Series C round, achieving a $1 billion valuation.
/ 19 / Hebbia: The AI Assistant for Financial and Legal Documents
Hebbia builds AI software for finance and law, offering advanced knowledge-work automation by organizing data and information for enterprises.
In professional services like finance, law, and consulting, the ability to efficiently retrieve information is critical. Many investigative processes require manual analysis of dense documentation.
A major pain point is information fragmentation caused by excessive SaaS usage within organizations. Finding relevant data often consumes significant time. A 2022 survey found that 47% of digital workers struggle to locate the information needed to perform their jobs effectively.
This is where Hebbia comes in. Founded to provide investment bankers and private equity analysts with neural search capabilities, it helps solve complex and ambiguous queries.
Its core product is Matrix. Users can upload any document type—PDFs, presentations, emails, images—and specify the questions they want answered. Hebbia then delivers detailed insights in a Microsoft Excel-like table format instantly.
In early 2025, Hebbia closed a $130 million Series B round led by Andreessen Horowitz, with participation from Index Ventures, Google Ventures, and Peter Thiel. The round brought its valuation to $700 million.
/ 20 / Mercor: What? HR Got Laid Off Too!
Mercor is an AI-driven recruitment and talent-matching platform that leverages AI throughout the hiring process to help companies find, screen, and pay employees more efficiently.
Mercor first uses AI to evaluate candidate profiles, followed by a ~20-minute AI video interview to assess suitability. It then quickly recommends top candidates matched to the company’s needs.
Beyond matching, Mercor manages payroll and ongoing performance tracking, streamlining the entire hiring and employment lifecycle for employers and talent alike.
Like other specialized AI companies, Mercor uses its own proprietary AI model, designed to mimic how recruiters evaluate resumes and conduct interviews.
Mercor has raised over $100 million in funding and achieved $75 million in ARR.
/ 21 / Pika: 5-Second Video + Photo = AI Magic That Puts Anyone in Any Video
Pika is an AI video generation platform developed by Pika Labs. Using AI, it can insert anyone or anything into any video.
Users simply upload a video longer than five seconds and a photo, then write a prompt to seamlessly integrate the photo into the video in a natural, convincing way—with no awkwardness.
Pika’s mission is to empower everyone to create fun short videos using AI. Its core strength lies in handling diverse creative styles—from 3D animation to cinematic sequences—while maintaining high visual quality at 1080p resolution.
After closing an $80 million Series B round in June 2024, Pika Labs reached a $470 million valuation, with reports suggesting a potential valuation as high as $700 million.
/ 22 / Suno: Music Lovers Without Instrument Skills, Rejoice!
Suno is an AI-powered music generation platform. Users can create realistic songs—including vocals and instruments—or purely instrumental tracks by simply entering text prompts.
Powered by its proprietary Chirp model, Suno meets diverse user needs. Importantly, Chirp does not copy existing songs, ensuring originality and uniqueness in AI-generated music.
Currently, Suno supports over 50 genres—including pop, rock, folk, and electronic—and can even generate niche styles like traditional Chinese folk music and opera. Its “Style Extend” feature lets users generate music in a similar style based on uploaded audio samples.
In 2024, the company raised $125 million, reaching a $500 million valuation. Its user base grew from 10 million to 25 million during the year.
/ 23 / StackBlitz: Building Software Is No Longer a Dream
StackBlitz is a browser-based integrated development environment (IDE) that allows developers to build, run, and deploy full-stack web applications directly from their web browsers.
Recognizing that few developers build web apps entirely within browsers, StackBlitz introduced WebContainers technology to make browser-native application development possible.
This opens up a new pathway for app development. Software creation is no longer exclusive to traditional developers—anyone can now create software without needing to learn programming or code.
In January 2025, AI coding company StackBlitz closed a new funding round exceeding $100 million. The company also announced its registered user count had surpassed 2 million.
/ 24 / Clay: The Precision Hunter for Sales Leads
Clay is a pre-sales lead development and management platform. By combining powerful AI with data from over 100 external sources—including public databases and social media—it helps businesses optimize marketing workflows and deliver actionable insights to sales and marketing teams.
Operating as a SaaS platform, Clay offers API integration and various data tools, automatically generating spreadsheets. From prospecting and outreach to retention, Clay drafts personalized outreach emails for each client, sparking interest and resonance.
In June 2024, Clay announced a $62 million funding round led by Meritech Capital, with participation from Sequoia Capital, First Round, BoxGroup, Boldstart, and several leaders in sales and marketing.
/ 25 / Captions: Taking On CapCut and Premiere Pro
Captions is an AI-powered video editing and content creation platform. It offers an all-in-one solution for creating and editing “talking-head” videos, streamlining the entire workflow from scriptwriting and recording to editing and publishing.
While popular tools like CapCut offer broad functionality, they’re not always optimal for specific use cases. Captions focuses on short-form talking-head videos, helping users fix issues like unnatural facial expressions and wandering eyes. Features include auto-captioning and AI-powered eye contact correction, helping novice speakers appear more natural.
Powered by its proprietary strong models, Captions can also generate AI-created talking avatars—heralded as ending the era of crude lip-syncing AI videos and ushering in a new generation of avatar technology.
The Mirage model is dubbed “the world’s first AI model designed specifically for UGC content generation.” Users input basic information to customize avatars—including appearance, personality, and clothing—and generate short-form UGC videos.
In July 2024, Captions closed a $60 million Series C round and was highlighted in an a16z report as one of the most profitable AI audio-video apps.
/ 26 / Decagon: Generative AI Agents for Enterprise Customer Support
Decagon builds AI agents for large enterprises, focusing on customer support and customer experience. Unlike rigid chatbots of the past, Decagon’s evolving AI Agent Engine truly understands customer needs and resolves issues.
Its core product, AI Agent Engine, uses generative AI to handle complex customer inquiries, deliver personalized responses, and seamlessly integrate with existing customer support systems.
The platform also supports multi-channel communication—including chat, email, and voice calls—enabling 24/7 service. Additionally, Decagon launched Decagon Voice to deliver high-fidelity voice customer service experiences.
/ 27 / OpenEvidence: The Doctor’s “AI Co-Pilot”
OpenEvidence is an AI software designed specifically for medical professionals. It aggregates the latest research from authoritative medical journals and guidelines to help clinicians make fast, accurate decisions during diagnosis and treatment.
Its built-in features are robust, including over 50 clinical calculators such as CHA2DS2-VASc, SOFA, and APACHE II, aiding risk assessment and scoring. It also supports various clinical scenarios—automatically generating patient handouts and discharge instructions, providing detailed drug information and usage guidance, and more.
Currently, OpenEvidence is free for U.S. medical professionals—users only need to verify their credentials. The platform is widely used by clinicians in over 7,000 healthcare institutions across 185 countries, becoming an indispensable tool in daily clinical practice.
As of early 2025, Decagon had raised $100 million in funding, including a $65 million Series B round led by Bain Capital Ventures, with participation from Accel and Andreessen Horowitz.
/ 28 / Vannevar Labs: Wait, Can AI Be a Spy Now?
Vannevar Labs applies AI to national defense and intelligence, using cutting-edge technology to protect national security.
Founded by former officials from the CIA, NSA, and Department of Defense, the company develops AI and machine learning technologies to address critical challenges in national security.
Decrypt is Vannevar Labs’ core product. It enables decentralized collection of global public data, making it difficult to trace data origins, while flexible querying allows users to quickly find, share, utilize, and distribute mission-relevant information.
As of January 2023, Vannevar Labs had closed a $75 million Series B round, bringing total funding above $90 million. In 2023, it achieved $49 million in annual revenue and became profitable within five years of founding.
/ 29 / Photoroom: The AI Background Remover—Effortless Image Editing
Photoroom is an AI-powered image editing software that provides fast, professional-grade tools for e-commerce, content creators, and small businesses—helping users quickly generate high-quality product images, social media covers, and profile pictures.
The software automatically detects and removes image backgrounds, highlighting the subject. Alternatively, it can generate realistic backgrounds from portrait photos. Users can also use AI to erase unwanted objects or people, greatly simplifying the editing process.
Photoroom emphasizes in-house model development, achieving image generation speeds 40% faster than other AI vision platforms.
In February 2024, the company raised $43 million in a Series B round at a $500 million valuation, led by Balderton Capital.
/ 30 / LangChain: Open Source Empowerment for Building Complex LLM Applications
LangChain is an open-source framework designed to help developers build applications powered by large language models (LLMs).
Through modular components, it simplifies interactions with LLMs and supports integration with external data sources, APIs, and services—enabling rapid development of complex AI applications.
LangChain’s goal is to enable LLMs not just to generate text, but also to interact with external systems and execute sophisticated application logic.
In early 2024, LangChain announced it had raised $25 million in a Series A round led by Sequoia Capital.
/ 31 / Midjourney: Image Generation at the Speed of Thought
Midjourney is an AI image generation service that excels at transforming text prompts into vivid visual images.
Users can generate high-resolution images from scratch by specifying parameters—covering landscapes, portraits, animals, abstract art, and more. It also supports style transfer, allowing users to apply the characteristics of one image onto another.
Additionally, Midjourney offers an intuitive, user-friendly interface that lets users experiment in real time with different inputs, styles, and settings. This interactive approach empowers artists, designers, and enthusiasts to actively engage in the creative process and explore endless artistic possibilities.
Midjourney boasts over 13 million users and $200 million in revenue—an astonishing achievement given it has raised zero outside funding!
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
