Top 100 AI startups in USA
Aug 06, 2026
|
Like
30
1
Funding: $925.4M
Etched.ai is an AI chip startup that develops Sohu, a chip designed specifically for running transformer models.
Etched.ai is an AI chip startup that develops Sohu, a chip designed specifically for running transformer models.
2
Funding: $21.7M
Dili focuses on AI for compliance in construction projects, particularly those getting some kind of federal funding. Given high stakes in these projects, Dili’s architecture intends to prevent any LLM-based fuzziness from sneaking into the final product. Contemporary AI models are only used in the company’s data layer, the engine for taking unstructured documents and translating them into structured data. From there, a deterministic system sorts the data according to the complex-but-static compliance rules. When it works right, a task that used to take a full day’s work can now be dispatched in a matter of minutes.
Dili focuses on AI for compliance in construction projects, particularly those getting some kind of federal funding. Given high stakes in these projects, Dili’s architecture intends to prevent any LLM-based fuzziness from sneaking into the final product. Contemporary AI models are only used in the company’s data layer, the engine for taking unstructured documents and translating them into structured data. From there, a deterministic system sorts the data according to the complex-but-static compliance rules. When it works right, a task that used to take a full day’s work can now be dispatched in a matter of minutes.
3
Funding: $22M
Smallest.ai builds the full-stack voice AI infrastructure behind real-time, human-like enterprise voice agents. The company wants to make speaking to an AI agent indistinguishable from talking to a human. To do so, it’s developing a small voice model designed to mimic how humans process information by listening, thinking, and speaking simultaneously. The startup’s model serves as a real-time intelligence layer that enables natural customer conversations on specific topics, with virtually zero response lag. But if the model encounters a subject outside its limited knowledge base, Smallest.ai hands off the query to a large foundational model
Smallest.ai builds the full-stack voice AI infrastructure behind real-time, human-like enterprise voice agents. The company wants to make speaking to an AI agent indistinguishable from talking to a human. To do so, it’s developing a small voice model designed to mimic how humans process information by listening, thinking, and speaking simultaneously. The startup’s model serves as a real-time intelligence layer that enables natural customer conversations on specific topics, with virtually zero response lag. But if the model encounters a subject outside its limited knowledge base, Smallest.ai hands off the query to a large foundational model
4
Funding: $300M
Simile is building a foundation model that predicts human behavior in any situation, and a product that deploys it at scale. It creates simulated AI users for areas like marketing and product research. The startup’s stated mission of simulating “all eight billion people on earth, accurately and honestly” is preposterous — the whole reason to conduct market research is because humans are unpredictable, guided, as we are, by both emotions and reason.
Simile is building a foundation model that predicts human behavior in any situation, and a product that deploys it at scale. It creates simulated AI users for areas like marketing and product research. The startup’s stated mission of simulating “all eight billion people on earth, accurately and honestly” is preposterous — the whole reason to conduct market research is because humans are unpredictable, guided, as we are, by both emotions and reason.
5
Funding: $15M
Infinity is building software to make it easier for AI chips to run AI models. It's s CUDA-alternative kernel software that works with any type of chip, like SRAM, GPUs, phone chips, and Systolic Arrays. Infinity is also building a universal inference library to run on all chips, allowing these chips to automate replicating state-of-the-art research results.
Infinity is building software to make it easier for AI chips to run AI models. It's s CUDA-alternative kernel software that works with any type of chip, like SRAM, GPUs, phone chips, and Systolic Arrays. Infinity is also building a universal inference library to run on all chips, allowing these chips to automate replicating state-of-the-art research results.
6
Funding: $10M
Mirai allows to deploy and run models of any architecture directly on user devices. It developed fastest inference engine built from scratch for Apple devices.
Mirai allows to deploy and run models of any architecture directly on user devices. It developed fastest inference engine built from scratch for Apple devices.
7
Funding: $42.4B
xAI is Elon Musk's me-too startup, which couldn't resist the AI craze and invested heavily in the project. Initially, the company created the AI chatbot Grok for X/Twitter users, and later grew into a universal LLM provider for users and companies (via API). Grok's distinguishing feature is its more censorship-free worldview (which has occasionally led to controversy). Grok also initially focused on real-time context generation from online data (rather than pre-training). In addition to language, Grok can generate images. xAI also runs the world's largest supercomputer, Colossus and is engaged in AGI research (accoring to Elon Mask, xAI seeks to understand the true nature of the universe).
xAI is Elon Musk's me-too startup, which couldn't resist the AI craze and invested heavily in the project. Initially, the company created the AI chatbot Grok for X/Twitter users, and later grew into a universal LLM provider for users and companies (via API). Grok's distinguishing feature is its more censorship-free worldview (which has occasionally led to controversy). Grok also initially focused on real-time context generation from online data (rather than pre-training). In addition to language, Grok can generate images. xAI also runs the world's largest supercomputer, Colossus and is engaged in AGI research (accoring to Elon Mask, xAI seeks to understand the true nature of the universe).
8
Funding: $2.5B
SambaNova Systems produces chips based on the Reconfigurable Dataflow Unit architecture, which allows more flexible distribution of computations and memory, adaptation to different types of models. This architecture supports the "composition of experts" mechanism - when several specialized models work on certain parts of the data/tasks. The chip has multi-level memory and is optimized for large open models - Llama, DeepSeek, etc. It uses low-precision formats to accelerate computations and strives to achieve an optimal ratio of performance / throughput per watt, that's why RDU is very energy efficient and enables more compact hardware infrastructure. The company offers a full-featured stack: not just hardware, but also a software platform, a cloud service and the ability to deploy an on-premises system. SambaNova Systems is focused on open technologies and standards.
SambaNova Systems produces chips based on the Reconfigurable Dataflow Unit architecture, which allows more flexible distribution of computations and memory, adaptation to different types of models. This architecture supports the "composition of experts" mechanism - when several specialized models work on certain parts of the data/tasks. The chip has multi-level memory and is optimized for large open models - Llama, DeepSeek, etc. It uses low-precision formats to accelerate computations and strives to achieve an optimal ratio of performance / throughput per watt, that's why RDU is very energy efficient and enables more compact hardware infrastructure. The company offers a full-featured stack: not just hardware, but also a software platform, a cloud service and the ability to deploy an on-premises system. SambaNova Systems is focused on open technologies and standards.
9
Funding: $2.4B
Groq creates hardware AI accelerators for large language models that improve performance and reduce power consumption compared to classic GPU accelerators. It produces a Language Processing Unit (LPU) chip that uses exclusively on-chip SRAM memory (without external DRAM/HBM modules), which requires installing multiple chips for scalable payloads. Its architecture minimizes elements associated with unpredictable behavior (branch prediction, caches, etc.). Groq enables best performance for small and medium batch/task sizes, especially if the model fits into their configuration. The software-defined, single-core architecture removes traditional software complexity while continuous, token-based execution delivers consistent performance without tradeoffs. Groq licenses its technology to NVidia. The company also provides its cloud AI inference platform built for developers - available in public, private, or co-cloud instances.
Groq creates hardware AI accelerators for large language models that improve performance and reduce power consumption compared to classic GPU accelerators. It produces a Language Processing Unit (LPU) chip that uses exclusively on-chip SRAM memory (without external DRAM/HBM modules), which requires installing multiple chips for scalable payloads. Its architecture minimizes elements associated with unpredictable behavior (branch prediction, caches, etc.). Groq enables best performance for small and medium batch/task sizes, especially if the model fits into their configuration. The software-defined, single-core architecture removes traditional software complexity while continuous, token-based execution delivers consistent performance without tradeoffs. Groq licenses its technology to NVidia. The company also provides its cloud AI inference platform built for developers - available in public, private, or co-cloud instances.
10
Funding: $781M
ElevenLabs develops voice AI models for content creators and publishers. It has created two platforms: the Agents Platform for improving customer engagement and Creative Platform for creating AI-powered voice content. The platform's capabilities include converting text into realistic speech in over 70 languages, configuring, deploying, and monitoring conversational agents, creating studio-quality tracks in any genre and style, transcribing any audio with the highest accuracy, and creating an exact copy of any voice. All platform capabilities are accessible via an API. ElevenLabs is used to create films, advertising, audiobooks and podcasts.
ElevenLabs develops voice AI models for content creators and publishers. It has created two platforms: the Agents Platform for improving customer engagement and Creative Platform for creating AI-powered voice content. The platform's capabilities include converting text into realistic speech in over 70 languages, configuring, deploying, and monitoring conversational agents, creating studio-quality tracks in any genre and style, transcribing any audio with the highest accuracy, and creating an exact copy of any voice. All platform capabilities are accessible via an API. ElevenLabs is used to create films, advertising, audiobooks and podcasts.
11
Funding: $337M
Odyssey is an AI lab focused on creating universal world models, which the company calls new form of audiovisual intelligence. These models are to form the basis of the next generation of games, films, education content, training simulations and advertising. These models don't just generate video - they enable interaction with the 3D world i.e. creation of interactive videos. The company achieves this through a new multi-stage training pipeline that transforms the model into a causal behavioral video model that reacts to actions in real time and continuously responds to input. As you play the video, you shape it in real time using natural text prompts, similar to communicating with a language model.
Odyssey is an AI lab focused on creating universal world models, which the company calls new form of audiovisual intelligence. These models are to form the basis of the next generation of games, films, education content, training simulations and advertising. These models don't just generate video - they enable interaction with the 3D world i.e. creation of interactive videos. The company achieves this through a new multi-stage training pipeline that transforms the model into a causal behavioral video model that reacts to actions in real time and continuously responds to input. As you play the video, you shape it in real time using natural text prompts, similar to communicating with a language model.
12
Funding: $327M
Fireworks AI provides cloud-based platform that enables developers to build, customize and scale AI applications using open-source models. It features a library of ready-made models and enables scaling inference at minimal cost. It also includes coding assistance tools (IDE assistants, code generation, debugging agents), agent systems for creating multi-stage reasoning, planning and execution pipelines, ready-to-use enterprise assistants (for summarization, semantic search, personalized recommendations), enterprise RAG search for knowledge bases. The company provides SDK for prototyping, quality assessment, and scaling with confidence.
Fireworks AI provides cloud-based platform that enables developers to build, customize and scale AI applications using open-source models. It features a library of ready-made models and enables scaling inference at minimal cost. It also includes coding assistance tools (IDE assistants, code generation, debugging agents), agent systems for creating multi-stage reasoning, planning and execution pipelines, ready-to-use enterprise assistants (for summarization, semantic search, personalized recommendations), enterprise RAG search for knowledge bases. The company provides SDK for prototyping, quality assessment, and scaling with confidence.
13
Funding: $160M
Resolve AI is developing an autonomous AI site reliability engineer (SRE) that automatically maintains software systems. It's a multi-agent system that connects to production development systems, code, services, infrastructure, telemetry and analyzes complex production issues in real time. The system learns from corporate operations manuals, wikis, chats and best practices. Resolve AI integrates with various monitoring, infrastructure tools via MCPs, APIs and webhooks. It's designed to meet stringent compliance standards, including SOC 2 Type II certification, GDPR and HIPAA.
Resolve AI is developing an autonomous AI site reliability engineer (SRE) that automatically maintains software systems. It's a multi-agent system that connects to production development systems, code, services, infrastructure, telemetry and analyzes complex production issues in real time. The system learns from corporate operations manuals, wikis, chats and best practices. Resolve AI integrates with various monitoring, infrastructure tools via MCPs, APIs and webhooks. It's designed to meet stringent compliance standards, including SOC 2 Type II certification, GDPR and HIPAA.
14
Funding: $160M
Loop provides a platform that manages logistics and supply chain data through automation and analytics tools. It's using AI to offer companies predictive remedies - takes unstructured data (PDFs with no optically recognized characters, sheets of paper, digital messages) and gives it structure, in order to automate tasks. Loop makes the automation possible by developing a harness that coordinates multiple AI models. This helps companies better identify where they may be losing money or time, or spot the risks of over- or under-supplying a given product.
Loop provides a platform that manages logistics and supply chain data through automation and analytics tools. It's using AI to offer companies predictive remedies - takes unstructured data (PDFs with no optically recognized characters, sheets of paper, digital messages) and gives it structure, in order to automate tasks. Loop makes the automation possible by developing a harness that coordinates multiple AI models. This helps companies better identify where they may be losing money or time, or spot the risks of over- or under-supplying a given product.
15
Funding: $150M
Inferact mission is to accelerate AI progress by making inference cheaper and faster.
Inferact mission is to accelerate AI progress by making inference cheaper and faster.
16
Funding: $130M
Scribe creates a platform that documents workflows across software applications. When someone completes a process or workflow, it produces a step-by-step guide using its browser extension and desktop app, along with text and screenshots. Those guides can be shared with colleagues or published in internal tools to prevent repeated questions, minimize errors, and accelerate onboarding. It also helps to uncover where automation and AI really produce returns — instead of becoming another sunk cost. It analyzes across workflows for what people are doing when they’re at work, and then it generalizes those up into being able to show you in a single view of glass, here are the actual workflows that are being done. Here’s how often, how long it takes, etc.
Scribe creates a platform that documents workflows across software applications. When someone completes a process or workflow, it produces a step-by-step guide using its browser extension and desktop app, along with text and screenshots. Those guides can be shared with colleagues or published in internal tools to prevent repeated questions, minimize errors, and accelerate onboarding. It also helps to uncover where automation and AI really produce returns — instead of becoming another sunk cost. It analyzes across workflows for what people are doing when they’re at work, and then it generalizes those up into being able to show you in a single view of glass, here are the actual workflows that are being done. Here’s how often, how long it takes, etc.
17
Funding: $100M
Emergent is an AI-powered platform for creating websites, web- and mobile apps. It lets you create and host projects, provides access to cutting-edge LLM models and chat interface where you can explain how the app should work in natural language, while the AI takes care of coding, design and deployment. As advertised, the generated code is ready to use and the resulting app is entirely yours. But Emergent is designed not only for complete beginners but also for experienced developers. You can create and deploy your own specialized agents tailored to your workflows, implement integrations that connect services, automate tasks and accelerate delivery, collaborate with your team in real time and sync projects with GitHub for version control.
Emergent is an AI-powered platform for creating websites, web- and mobile apps. It lets you create and host projects, provides access to cutting-edge LLM models and chat interface where you can explain how the app should work in natural language, while the AI takes care of coding, design and deployment. As advertised, the generated code is ready to use and the resulting app is entirely yours. But Emergent is designed not only for complete beginners but also for experienced developers. You can create and deploy your own specialized agents tailored to your workflows, implement integrations that connect services, automate tasks and accelerate delivery, collaborate with your team in real time and sync projects with GitHub for version control.
18
Funding: $85.5M
WitnessAI develops a security system that creates the confidence layer for enterprise AI. It enables to monitor all AI activities within organization, protect models and applications (using next-generation AI Firewall and content controls), detect shadow AI use, catalog entire AI inventory (applications, MCP servers, agents), visualize all conversations with AI chats, including prompts and responses, implement intelligent controls and advanced business rules that record AI interactions among employees and agents, adapt to different usage contexts and reduce the cost of AI compliance.
WitnessAI develops a security system that creates the confidence layer for enterprise AI. It enables to monitor all AI activities within organization, protect models and applications (using next-generation AI Firewall and content controls), detect shadow AI use, catalog entire AI inventory (applications, MCP servers, agents), visualize all conversations with AI chats, including prompts and responses, implement intelligent controls and advanced business rules that record AI interactions among employees and agents, adapt to different usage contexts and reduce the cost of AI compliance.
19
Funding: $60M
Outtake is a digital trust platform automating threat classification, detection and response to enhance enterprise cybersecurity measures for AI-driven attack age. Its agentic AI secures modern attack surfaces with advanced search, real-time threat classification, and automated response. AI agents adapt and improve as fast as the threats themselves, learning and responding to new attack patterns in real-time. Outtake names among its customers OpenAI, Pershing Square, AppLovin, and federal agencies.
Outtake is a digital trust platform automating threat classification, detection and response to enhance enterprise cybersecurity measures for AI-driven attack age. Its agentic AI secures modern attack surfaces with advanced search, real-time threat classification, and automated response. AI agents adapt and improve as fast as the threats themselves, learning and responding to new attack patterns in real-time. Outtake names among its customers OpenAI, Pershing Square, AppLovin, and federal agencies.
20
Funding: $50M
Rox leverages AI agents to enhance sales productivity, providing actionable insights that help close deals.
Rox leverages AI agents to enhance sales productivity, providing actionable insights that help close deals.
21
Funding: $40M
Articul8 develops specialized GenAI systems that operate within clients' own local IT environment, rather than relying on cloud general-purpose models. Instead of selling standalone models, the company offers its technology as software applications and AI agents tailored to specific business functions, targeting regulated industries such as energy, manufacturing, aerospace, financial services, semiconductors, where data accuracy, auditability and control are critical. The company develops agentic reasoning engine that autonomously assembles the right squad of domain-specific agents for every mission.
Articul8 develops specialized GenAI systems that operate within clients' own local IT environment, rather than relying on cloud general-purpose models. Instead of selling standalone models, the company offers its technology as software applications and AI agents tailored to specific business functions, targeting regulated industries such as energy, manufacturing, aerospace, financial services, semiconductors, where data accuracy, auditability and control are critical. The company develops agentic reasoning engine that autonomously assembles the right squad of domain-specific agents for every mission.
22
Funding: $40M
NeoCognition is a research lab developing general-purpose AI agents capable of self-learning and becoming experts in any field, similar to how humans learn. Their learning process is essentially the process of constructing a model of the world for any profession (any given microcosm). NeoCognition plans to sell its agent systems primarily to enterprises, including established SaaS companies, who can use them to create agent workers or improve existing products.
NeoCognition is a research lab developing general-purpose AI agents capable of self-learning and becoming experts in any field, similar to how humans learn. Their learning process is essentially the process of constructing a model of the world for any profession (any given microcosm). NeoCognition plans to sell its agent systems primarily to enterprises, including established SaaS companies, who can use them to create agent workers or improve existing products.
23
Funding: $36M
Sandbar creates AI smart ring focused on note-taking. It has a microphone that can be activated using a flat, touch-sensitive panel at the top. You can hold this touch panel to record notes, chat with an AI assistant on the accompanying phone app, and access media controls like play, pause, skip tracks, and control the volume. Notes are saved to phone app that can be used on its own to take notes in case the ring is charging or has been misplaced. The startup plans to start shipping the smart ring in 2026.
Sandbar creates AI smart ring focused on note-taking. It has a microphone that can be activated using a flat, touch-sensitive panel at the top. You can hold this touch panel to record notes, chat with an AI assistant on the accompanying phone app, and access media controls like play, pause, skip tracks, and control the volume. Notes are saved to phone app that can be used on its own to take notes in case the ring is charging or has been misplaced. The startup plans to start shipping the smart ring in 2026.
24
Funding: $34.5M
InsightFinder specializes in AI-powered observability and predictive analytics for IT operations.
InsightFinder specializes in AI-powered observability and predictive analytics for IT operations.
25
Funding: $27.9M
Limitless AI is developing wearable device with a personal AI assistant that helps you capture, remember, and utilize information. It's a small pendant that attaches to shirt like a wireless microphone or wears as a necklace, and allows to record conversations. Its primary purpose is to record and transcribe meetings, eliminating the need to take notes. The device is waterproof, has a 100-hour battery life and charges via USB-C port. It also features a "consent mode" that prevents recording unless the other person has explicitly consented. The companion app is free and includes unlimited audio storage and 10 hours per month of AI features such as transcription, summary creation, and note-taking. Acquired by Meta
Limitless AI is developing wearable device with a personal AI assistant that helps you capture, remember, and utilize information. It's a small pendant that attaches to shirt like a wireless microphone or wears as a necklace, and allows to record conversations. Its primary purpose is to record and transcribe meetings, eliminating the need to take notes. The device is waterproof, has a 100-hour battery life and charges via USB-C port. It also features a "consent mode" that prevents recording unless the other person has explicitly consented. The companion app is free and includes unlimited audio storage and 10 hours per month of AI features such as transcription, summary creation, and note-taking. Acquired by Meta
26
Funding: $27M
CopilotKit develops framework for creating interactive in-app AI agents. The company believes that in the future, agents will run inside apps, understand user actions, perform necessary actions, and display useful interfaces instead of simply returning long blocks of text. The company's open AG-UI protocol standardizes how AI agents connect and interact with user interfaces (such as a web browser or app), enabling such features as threaded chat, interface tool calls, and state exchange to enable human-assisted functionality. CopilotKit is also developing an enterprise toolkit based on AG-UI, adding support, self-service deployment features, and other necessary features for companies planning to integrate agents into their products.
CopilotKit develops framework for creating interactive in-app AI agents. The company believes that in the future, agents will run inside apps, understand user actions, perform necessary actions, and display useful interfaces instead of simply returning long blocks of text. The company's open AG-UI protocol standardizes how AI agents connect and interact with user interfaces (such as a web browser or app), enabling such features as threaded chat, interface tool calls, and state exchange to enable human-assisted functionality. CopilotKit is also developing an enterprise toolkit based on AG-UI, adding support, self-service deployment features, and other necessary features for companies planning to integrate agents into their products.
27
Funding: $25M
Deccan AI specializes in providing high-quality, human-labeled data for AI model training and evaluation.
Deccan AI specializes in providing high-quality, human-labeled data for AI model training and evaluation.
28
Funding: $22M
Runpod is a cloud platform designed for GPUs, enabling developers to deploy customized full-stack AI applications.
Runpod is a cloud platform designed for GPUs, enabling developers to deploy customized full-stack AI applications.
29
Funding: $20.1M
Vapi is a voice AI platform that allows developers to build, test and deploy voice agents for phone calls and customer support.
Vapi is a voice AI platform that allows developers to build, test and deploy voice agents for phone calls and customer support.
30
Funding: $18.2M
Shade is developing a cloud-based media storage platform designed for agencies, sports media teams, and podcasters to easily store and search their media using natural language queries. The startup claims that their search doesn't just display a specific video but pinpoints the exact moment in the video where the scene corresponding to the search query occurs. The tool also automatically transcribes the audio for easier searching. Users can search by meaning, transcript, and facial recognition for tagged individuals. For example, users can search for "person with laptop in snow," and the system will return all relevant clips with timestamps.
Shade is developing a cloud-based media storage platform designed for agencies, sports media teams, and podcasters to easily store and search their media using natural language queries. The startup claims that their search doesn't just display a specific video but pinpoints the exact moment in the video where the scene corresponding to the search query occurs. The tool also automatically transcribes the audio for easier searching. Users can search by meaning, transcript, and facial recognition for tagged individuals. For example, users can search for "person with laptop in snow," and the system will return all relevant clips with timestamps.
31
Funding: $10.5M
Risotto is an AI ChatOps for IT support, boosting efficiency and security with instant resolution and 24/7 automation, all via chat.
Risotto is an AI ChatOps for IT support, boosting efficiency and security with instant resolution and 24/7 automation, all via chat.
32
Funding: $11M
Littlebird is the only full-context AI. It works in the background, observing your screen and transcribing your meetings, to build a private memory of your work.
Littlebird is the only full-context AI. It works in the background, observing your screen and transcribing your meetings, to build a private memory of your work.
33
Funding: $9.5M
Guide Labs is a software development company that builds interpretable models to explain reasoning that are easy to steer, debug, and align.
Guide Labs is a software development company that builds interpretable models to explain reasoning that are easy to steer, debug, and align.
34
Funding: $9.3M
Onton is developing AI-chat-based search engine for online furniture stores. The company uses so-called neurosymbolic architecture to overcome the hallucination problems inherent in LLM and provide higher-quality, logical search results. The startup's model can also be trained using real-world information that doesn't necessarily need to be included in product descriptions. For example, if you search for pet-friendly furniture, the model will learn and start return only furniture that is stain- and scratch-resistant. You can also upload an image to Onton's chat to generate a desired design for your home or office, and Onton will then select furniture based on it. The service also offers an infinite image generation canvas, where you can add existing images along with products you've found for inspiration.
Onton is developing AI-chat-based search engine for online furniture stores. The company uses so-called neurosymbolic architecture to overcome the hallucination problems inherent in LLM and provide higher-quality, logical search results. The startup's model can also be trained using real-world information that doesn't necessarily need to be included in product descriptions. For example, if you search for pet-friendly furniture, the model will learn and start return only furniture that is stain- and scratch-resistant. You can also upload an image to Onton's chat to generate a desired design for your home or office, and Onton will then select furniture based on it. The service also offers an infinite image generation canvas, where you can add existing images along with products you've found for inspiration.
35
Funding: $9M
Probably develops technology that can prevent LLM hallucinations and simple factual errors. It's a sophisticated security system, which the startup describes as a "mechanical data suit." LLM results are verified using a deterministic validation system that returns any results that don't match the dataset. Each result is accompanied by a link to the source and an audit trail of the development process, a practice that is becoming increasingly common among AI tools.
Probably develops technology that can prevent LLM hallucinations and simple factual errors. It's a sophisticated security system, which the startup describes as a "mechanical data suit." LLM results are verified using a deterministic validation system that returns any results that don't match the dataset. Each result is accompanied by a link to the source and an audit trail of the development process, a practice that is becoming increasingly common among AI tools.
36
Funding: $8.4M
Nomadic helps companies developing self-driving cars, robots manipulating the physical environment, or autonomous construction equipment to analyse and organize collected video data for evaluation and training. Becuase the most valuable data depicts events that rarely occur and can befuddle inexperienced physical AI models, Nomadic has developed a platform that turns footage into a structured, searchable dataset through a collection of vision language models. That, in turn, allows for better fleet monitoring and the creation of unique datasets for reinforcement learning and faster iteration. Customers like Zoox, Mitsubishi Electric, Natix Network, and Zendar are already using the platform to develop intelligent machines.
Nomadic helps companies developing self-driving cars, robots manipulating the physical environment, or autonomous construction equipment to analyse and organize collected video data for evaluation and training. Becuase the most valuable data depicts events that rarely occur and can befuddle inexperienced physical AI models, Nomadic has developed a platform that turns footage into a structured, searchable dataset through a collection of vision language models. That, in turn, allows for better fleet monitoring and the creation of unique datasets for reinforcement learning and faster iteration. Customers like Zoox, Mitsubishi Electric, Natix Network, and Zendar are already using the platform to develop intelligent machines.
37
Funding: $7M
Atomic Canyon uses artificial intelligence to improve efficiency and optmize operations of nuclear power plants. The company is developing Neutron, an AI search engine for searching billions of pages of nuclear documentation. Using the Nuclear Regulatory Commission (NRC) database and the world's fastest supercomputer at Oak Ridge National Laboratory (ORNL), it understands and processes nuclear terminology with unprecedented accuracy, interpreting complex documents to let executives make more informed data-driven decisions. Neutron is based on open-source AI model FERMI intended for sentence vector representations, specifically designed for nuclear data. Atomic Canyon is working in partnership with PG&E's Diablo Canyon power plant, the only remaining nuclear power plant in California.
Atomic Canyon uses artificial intelligence to improve efficiency and optmize operations of nuclear power plants. The company is developing Neutron, an AI search engine for searching billions of pages of nuclear documentation. Using the Nuclear Regulatory Commission (NRC) database and the world's fastest supercomputer at Oak Ridge National Laboratory (ORNL), it understands and processes nuclear terminology with unprecedented accuracy, interpreting complex documents to let executives make more informed data-driven decisions. Neutron is based on open-source AI model FERMI intended for sentence vector representations, specifically designed for nuclear data. Atomic Canyon is working in partnership with PG&E's Diablo Canyon power plant, the only remaining nuclear power plant in California.
38
Funding: $6M
AgentMail provides an API-first email platform that enables AI agents to send, receive, and manage emails with automation and analytics.
AgentMail provides an API-first email platform that enables AI agents to send, receive, and manage emails with automation and analytics.
39
Funding: $5.8M
Luminal develops a compiler for optimizing machine learning models and provides cloud platform for running optimized models. Companies upload their Huggingface models and their weights to the Luminal cloud and receive a serverless endpoint (i.e., you simply send a request for example, an image, text, or audio to a special URL and receive the result). Luminal compiles models into GPU code with zero overhead. Optimization methods allow to squeeze more computing power out of existing infrastructure. The compiler, which sits between the written code and the GPU hardware, effectively competes with Nvidia's proprietary CUDA stack.
Luminal develops a compiler for optimizing machine learning models and provides cloud platform for running optimized models. Companies upload their Huggingface models and their weights to the Luminal cloud and receive a serverless endpoint (i.e., you simply send a request for example, an image, text, or audio to a special URL and receive the result). Luminal compiles models into GPU code with zero overhead. Optimization methods allow to squeeze more computing power out of existing infrastructure. The compiler, which sits between the written code and the GPU hardware, effectively competes with Nvidia's proprietary CUDA stack.
40
Funding: $5.5M
VoiceRun is a platform for developing voice-enabled AI agents. Unlike visual agent-builders, it allows to program voice agent behavior using code, which gives greater flexibility. In addition to creating code-based agents, VoiceRun also allows users to conduct A/B testing and instantly deploy solutions with a single click through the VoiceRun cloud telephony system and swap models instantly, backed by enterprise-grade security. The company is focused on enterprise developers, helping companies, for example, implement AI in their customer support services or assist tech companies in launching voice-based products. For example, the platform is used to create an AI concierge for restaurant reservations.
VoiceRun is a platform for developing voice-enabled AI agents. Unlike visual agent-builders, it allows to program voice agent behavior using code, which gives greater flexibility. In addition to creating code-based agents, VoiceRun also allows users to conduct A/B testing and instantly deploy solutions with a single click through the VoiceRun cloud telephony system and swap models instantly, backed by enterprise-grade security. The company is focused on enterprise developers, helping companies, for example, implement AI in their customer support services or assist tech companies in launching voice-based products. For example, the platform is used to create an AI concierge for restaurant reservations.
41
Funding: $3.5M
Modelence is a Software Development firm offering typeScript cloud services AI apps production.
Modelence is a Software Development firm offering typeScript cloud services AI apps production.
42
Funding: $2.5M
First Voyage is developing a mobile app, Momo Self Care, featuring a virtual AI pet named Momo, who helps users to form desired habits. You care for Momo and it reminds you to complete habit-building tasks and rewards you with coins for completing them. The coins can be used to purchase in-app items to further personalize your pet. Users can also talk to Momo about self-care and the AI companion will recommend habits and tasks based on what you want to achieve. The most popular habits (according to the developers) are related to productivity, spirituality and mindfulness. Momo includes safety mechanisms, such as suggestion filters, to ensure that communication between the AI and users remains within acceptable guidelines.
First Voyage is developing a mobile app, Momo Self Care, featuring a virtual AI pet named Momo, who helps users to form desired habits. You care for Momo and it reminds you to complete habit-building tasks and rewards you with coins for completing them. The coins can be used to purchase in-app items to further personalize your pet. Users can also talk to Momo about self-care and the AI companion will recommend habits and tasks based on what you want to achieve. The most popular habits (according to the developers) are related to productivity, spirituality and mindfulness. Momo includes safety mechanisms, such as suggestion filters, to ensure that communication between the AI and users remains within acceptable guidelines.
43
Pebble makes a smart ring with artificial intelligence for making short notes. You press a button on the ring, speak your thought and it's sent to your phone, added to your notebook app and can be set as a reminder. The AI is enabled only in the Pebble app on the phone, which uses an open-source model that runs locally and doesn't send data to the cloud, so it's private and doesn't require an internet connection or subscription. The ring is made of durable stainless steel and is water-resistant. As advertised, the battery lasts for years - you never need to charge it.
44
Numenta is a startup by Jeff Hawkins, the author of renowned books on AI ("On Intelligence" and "A Thousand Brains: A New Theory of Intelligence") and his own theory of brain structure. Therefore, the startup's primary goal was initially to understand the neocortex (the primary cerebral cortex) and apply its principles to the creation of machine intelligence. The theory, called Hierarchical Temporal Memory (HTM) describes a learning and prediction model inspired by the cerebral cortex. The company has published open-source code of its framework (Thousand Brains Project) - a platform for sensory-motor learning. The company's main product, the Numenta Platform for Intelligent Computing (NuPIC), is also based on algorithms, data structures and architecture inspired by the brain.
45
Funding: $189B
OpenAI develops generative AI models: GPT for language, DALLE for images and Codex for code. The company's main product is the chatbot ChatGPT, which is used as a personal smart assistant and in enterprise customer support and knowledge management systems via APIs. The company also conducts scientific research aimed at achieving artificial general intelligence. The company's initial goal was the security and openness of artificial intelligence, but it has gradually moved toward closing and commercializing its ML and NLP technologies. According to CEO Sam Altman, AGI is achieving a certain amount of revenue. The primary OpenAI's investor is Microsoft and it has strategic partnership with NVidia. OpenAI is also a primary AI provider for the US Department of War
OpenAI develops generative AI models: GPT for language, DALLE for images and Codex for code. The company's main product is the chatbot ChatGPT, which is used as a personal smart assistant and in enterprise customer support and knowledge management systems via APIs. The company also conducts scientific research aimed at achieving artificial general intelligence. The company's initial goal was the security and openness of artificial intelligence, but it has gradually moved toward closing and commercializing its ML and NLP technologies. According to CEO Sam Altman, AGI is achieving a certain amount of revenue. The primary OpenAI's investor is Microsoft and it has strategic partnership with NVidia. OpenAI is also a primary AI provider for the US Department of War
46
Funding: $128.7B
Anthropic was founded by a group of former OpenAI employees who wanted to preserve the goal of AI safety. Anthropic is one of technology leaders in the AI market. Its core product is a family of LLM models under the Claude brand, which are used in the form of a chatbot and API for business. The strength of this model is quality code generation. The company promotes an approach called "Constitutional AI" according to which models are trained according to a set of principles, values, and rules (like Asimov's Laws of Robotics). In addition to selling commercial products, the company is actively researching the interpretability of AI models and AGI.
Anthropic was founded by a group of former OpenAI employees who wanted to preserve the goal of AI safety. Anthropic is one of technology leaders in the AI market. Its core product is a family of LLM models under the Claude brand, which are used in the form of a chatbot and API for business. The strength of this model is quality code generation. The company promotes an approach called "Constitutional AI" according to which models are trained according to a set of principles, values, and rules (like Asimov's Laws of Robotics). In addition to selling commercial products, the company is actively researching the interpretability of AI models and AGI.
47
Funding: $4.7B
Cerebras is building Wafer-Scale Engine (WSE) – the largest chip ever built for deep learning systems. The chip has a size of a silicon wafer, with a very large number of transistors and cores. Most of the memory on the chip is SRAM, with no or minimal (compared to SambaNova) external memory such as DRAM / HBM. This creates certain limitations in terms of flexibility of use and scaling of models, especially when the model is very large. But it has a very high bandwidth of the internal bus that enables super-fast data transfer between components inside one large plate. Each Cerebras system uses several such chips and requires significant cooling capacity. The WSE-chip powers the Cerebras CS-X - the AI supercomputer that enables less networking and a smaller footprint than a GPU-based cluster, and eliminate programming complexity by interacting with a single logical device at every scale. Cerebras also provides cloud learning/inference service for LLM companies like OpenAI.
Cerebras is building Wafer-Scale Engine (WSE) – the largest chip ever built for deep learning systems. The chip has a size of a silicon wafer, with a very large number of transistors and cores. Most of the memory on the chip is SRAM, with no or minimal (compared to SambaNova) external memory such as DRAM / HBM. This creates certain limitations in terms of flexibility of use and scaling of models, especially when the model is very large. But it has a very high bandwidth of the internal bus that enables super-fast data transfer between components inside one large plate. Each Cerebras system uses several such chips and requires significant cooling capacity. The WSE-chip powers the Cerebras CS-X - the AI supercomputer that enables less networking and a smaller footprint than a GPU-based cluster, and eliminate programming complexity by interacting with a single logical device at every scale. Cerebras also provides cloud learning/inference service for LLM companies like OpenAI.
48
Funding: $3.2B
Lambda is building a network of data centers focused on AI workloads. The company uses a proprietary design for rapidly deploying these high-density, high-cooling AI factories to meet modern AI requirements and maximize the performance per watt. Lambda's AI factories are modular: new GPUs, interconnects and cooling slots are integrated as they are delivered. Companies can lease capacity in these AI factories for model training and inference. Lambda also provides its own software for orchestrating AI workloads and a stack of software packages widely used in machine and deep learning. Each package is tested for compatibility with all Lambda's systems.
Lambda is building a network of data centers focused on AI workloads. The company uses a proprietary design for rapidly deploying these high-density, high-cooling AI factories to meet modern AI requirements and maximize the performance per watt. Lambda's AI factories are modular: new GPUs, interconnects and cooling slots are integrated as they are delivered. Companies can lease capacity in these AI factories for model training and inference. Lambda also provides its own software for orchestrating AI workloads and a stack of software packages widely used in machine and deep learning. Each package is tested for compatibility with all Lambda's systems.
49
Funding: $1.6B
Sierra develops a platform for creating, managing, and optimizing enterprise AI agents. Its agents accurately simulate the nuances of human interaction, enhance customer satisfaction, and always adhere to the brand style. Agent Studio allows for easy customization of customer scenarios without coding. The company also provides an Agent SDK for developers, enabling the rapid creation of agents using pre-built composable skills, fine-tuning capabilities, and advanced developer tools. Agents support voice recognition and speech generation and are suitable for automated phone support. The platform also assists live agents by providing real-time recommendations and automatically prepared responses.
Sierra develops a platform for creating, managing, and optimizing enterprise AI agents. Its agents accurately simulate the nuances of human interaction, enhance customer satisfaction, and always adhere to the brand style. Agent Studio allows for easy customization of customer scenarios without coding. The company also provides an Agent SDK for developers, enabling the rapid creation of agents using pre-built composable skills, fine-tuning capabilities, and advanced developer tools. Agents support voice recognition and speech generation and are suitable for automated phone support. The platform also assists live agents by providing real-time recommendations and automatically prepared responses.
50
Funding: $1.2B
World Labs develops spatially intelligent AI world models that are capable of perceiving, generating, reasoning and interacting with 3D world, unlocking the full potential of AI. The company believes that spatial intelligence will unlock new forms of storytelling, creativity, design, modeling and immersive experiences in both virtual and physical worlds. Its first product, Marble - is based on best-in-class generative 3D world models and enables anyone to create spatially consistent, highly accurate and robust 3D worlds using just a single image, video or text prompt. World Labs' founders include AI pioneer Fei-Fei Li and other world-renowned experts in machine learning, generative AI and computer vision.
World Labs develops spatially intelligent AI world models that are capable of perceiving, generating, reasoning and interacting with 3D world, unlocking the full potential of AI. The company believes that spatial intelligence will unlock new forms of storytelling, creativity, design, modeling and immersive experiences in both virtual and physical worlds. Its first product, Marble - is based on best-in-class generative 3D world models and enables anyone to create spatially consistent, highly accurate and robust 3D worlds using just a single image, video or text prompt. World Labs' founders include AI pioneer Fei-Fei Li and other world-renowned experts in machine learning, generative AI and computer vision.
