{
  "entities": [
    {
      "content": "12 Feb 2026 — Groq has told investors that it is projecting $500M in revenue for the year 2025,. $17 billion in cash payments across three installments by ...",
      "favicon": "https://sacra.com/static/img/favicon.ico",
      "published_date": "2026-08-05T04:10:25.0000000",
      "score": 0.5706919,
      "text": "Groq revenue, valuation & funding | Sacra Sacra Logo Features Customers Docs Pricing Unlock Full Access Sign In Your Feed Workspace Charts Highlights Questions Explore Companies Categories Reports About Features Customers Docs Pricing Community New Chat Your Feed Workspace Charts Highlights Questions Explore Companies Categories Reports Community Ask a question Get access to all our reports for $50/mo Read about the latest in the private markets and join a growing community. Valuation Model Expert Interviews Founders, funding Get Access Home > Companies > Groq Groq Follow AI inference chip and cloud tool for running large language models efficiently #ai #ai-chips Revenue $90.00M 2024 Valuation $3.60B 2025 Funding $901.13M 2024 View PDF Embed PDF View Model Details Headquarters Mountain View, United States CEO Jonathan Ross Website groq.com Milestones FOUNDING YEAR 2016 Listed In #ai #ai-chips Overview Reports Filings Highlight Unhighlight Revenue Sacra estimates that Groq generated $90M in revenue in 2024. The company primarily generates revenue from selling cloud services for companies to run AI on its chips, similar to how companies purchase access to OpenAI models or AI models from Amazon Web Services. Groq has told investors that it is projecting $500M in revenue for the year 2025,. Groq also sells its chip systems and data center operating services to other companies, including telecommunications company Bell Canada. The company reports that nearly 2 million developers and teams have used its services, indicating strong adoption of its GroqCloud platform. The revenue growth reflects the company's transition from a pure hardware play to a cloud-first business model, capitalizing on the massive demand for high-speed AI inference capabilities across enterprise customers and the developer community. Valuation & Funding Groq raised $750 million in September 2024, led by Disruptive, with participation from BlackRock and Neuberger Berman. This round valued the company at approximately $6.9 billion. Previous rounds included a $640 million Series D in August 2024 led by BlackRock Private Equity Partners, and a $300 million Series C in April 2021 co-led by Tiger Global Management and D1 Capital. The company raised over $1 billion in total funding since its founding in 2016. On December 24, 2025, Groq entered a non-exclusive licensing agreement with Nvidia Corp. for its inference technology, structured to deliver $17 billion in cash payments across three installments by the end of 2026. In February 2026, Groq distributed $7.6 billion to shareholders as the first major payout under this agreement, equal to approximately $64 per share, representing roughly three-quarters of the company's stock ownership. Shareholders are expected to receive an additional $2.7 billion when Groq redeems remaining shares, though timing has not been disclosed. The deal has been characterized by some industry observers as having a roughly $20 billion \"headline\" figure, though elements beyond the $17 billion cash licensing payments remain unclear. Key strategic investors across the company's funding history include Founders Fund, Addition, Neuberger Berman, Type One Ventures, Cisco Investments, Global Brain, Samsung Catalyst Fund, and The Spruce House Partnership. Product Groq is a specialized AI inference company built around its proprietary Language Processing Unit (LPU) chip architecture. The LPU uses a statically-scheduled tensor streaming processor design that moves data through a deterministic pipeline, delivering ultra-low latency and up to 10x energy efficiency compared to traditional GPUs. The core product is GroqCloud, a fully managed inference platform that developers can access through an OpenAI-compatible REST API. Developers can switch to Groq by changing just three lines of code—their API key, base URL, and model name—and immediately start streaming tokens at hundreds of tokens per second with sub-10 millisecond first-token latency. GroqCloud supports curated open-source models like Llama-3, Qwen, and Mixtral, with performance benchmarks showing speeds like 1,345 tokens per second for Llama-3 8B and 662 tokens per second for Qwen-3 32B. The platform includes Python and JavaScript SDKs, streaming responses, and integrations with popular frameworks like LangChain, LlamaIndex, and Vercel AI SDK. For enterprise customers, Groq offers GroqRack compute clusters—on-premises or colocation racks containing 64 to 576+ LPUs per rack. These systems target hyperscalers, sovereign clouds, and regulated industries requiring data residency, with customers like Aramco Digital ordering hundreds of racks for large-scale deployments. Business Model Groq operates a three-tier business model that ramps customers from cloud services to dedicated hardware. The company's B2B go-to-market approach targets both individual developers and enterprise customers through different engagement models. GroqCloud uses a pay-per-token pricing model similar to OpenAI, where customers pay only for tokens generated. The platform offers both self-serve access with no minimum commitments and enterprise plans with annual volume commitments and volume discounts. This usage-based model allows customers to easily estimate costs before engaging with sales teams. The hardware business involves selling GroqRack systems directly to enterprises, telecommunications companies, and cloud providers. These customers typically require on-premises deployment for data sovereignty, regulatory compliance, or performance requirements that cloud services cannot meet. Groq's cost structure benefits from its vertically integrated approach, controlling both chip design and the software stack. The company manufactures chips through foundry partnerships, currently using GlobalFoundries' 14nm process with plans to move to Samsung's 4nm process for next-generation chips. This vertical integration allows Groq to optimize the entire inference pipeline while maintaining higher margins than pure software plays. Groq has added a technology licensing revenue stream through a non-exclusive agreement with Nvidia, enabling Nvidia to integrate Groq's inference technology while Groq continues operating independently under new CEO Simon Edwards. Founder Jonathan Ross, President Sunny Madra, and other team members joined Nvidia to support scaling the licensed technology, while GroqCloud services continue without interruption. This licensing model represents a fourth revenue stream alongside cloud services, hardware sales, and enterprise contracts. Competition GPU incumbents Nvidia dominates the AI inference market with its H200, B200, and GH200 Grace-Hopper systems, maintaining over 80% of deployed inference GPUs. While Nvidia typically requires 8-16 GPUs to match Groq's token-per-second performance on large language models, the company's CUDA software ecosystem remains the stickiest competitive barrier. Nvidia is pushing its Blackwell B200 and GB200 systems to reset the performance-per-watt curve while bundling networking components to lock in customers. Nvidia's non-exclusive licensing agreement with Groq for its inference technology represents a strategic shift from pure competition to technology partnership, enabling Nvidia to integrate Groq's specialized inference capabilities into its offerings. This arrangement allows Nvidia to strengthen its inference portfolio while Groq continues operating its own cloud and hardware businesses independently. AMD is positioning its MI300X, MI325X, and upcoming MI355X chips as cost-effective alternatives, with some benchmarks showing advantages on specific workloads. However, AMD's ROCm tooling still lags significantly behind CUDA in developer adoption, and the company faces supply chain challenges that have delayed competitive responses to Nvidia's latest generations. Intel's Gaudi 3 chips claim 50% better performance than H100 at 40% lower total cost of ownership, but software fragmentation across oneAPI and SynapseAI has slowed enterprise adoption. Intel is targeting cloud partnerships by bundling Gaudi 3 with other data center components. Specialized inference chips Cerebras and SambaNova are building competing ASIC architectures optimized for AI inference, arguing that GPU economics break down when batch sizes are small for applications like chat, agent loops, and code assistance. These companies are positioning around extreme throughput-per-watt and deterministic latency, similar to Groq's approach. Hyperscalers are developing internal silicon solutions including Google's TPU v5, Amazon's Inferentia 3, and Microsoft's Cobalt and Maia chips. While these primarily serve internal workloads, they represent potential competition for third-party inference demand as hyperscalers may offer these capabilities to external customers. Cloud inference platforms Traditional cloud providers like AWS, Google Cloud, and Microsoft Azure are expanding their AI inference offerings, leveraging existing customer relationships and integrated billing. These platforms can bundle inference with other cloud services, creating switching costs that pure-play inference companies like Groq must overcome through superior performance or pricing. TAM Expansion New products and technology Groq is developing next-generation 4nm LPU chips through its partnership with Samsung Foundry, promising significant performance-per-watt improvements over current 14nm parts. This technology advancement opens opportunities in higher-value systems markets and enables support for larger context windows and real-time multimodal models. The company's acquisition of Definitive Intelligence bundled data preparation, orchestration, and dashboard capabilities with Groq hardware, allowing enterprises to purchase turnkey inference systems rather than just chips. This vertical software stack expansion moves Groq up the value chain from hardware provider to complete solution vendor. GroqCloud's tokens-as-a-service model transforms raw silicon into recurring software revenue streams. The Series D funding will add over 100,000 LPUs to the cloud by early 2025, enabling support for more sophisticated AI applications and larger-scale deployments. Customer base expansion Groq has grown from zero to over 360,000 developers in 18 months, with 75% of Fortune 100 companies maintaining accounts on the platform. By supporting open-source communities around models like Llama 3, Gemma, and Mixtral, Groq expands beyond Big Tech buyers into small and medium business developer tool budgets. The company's reseller partnership with Carahsoft and FedRAMP roadmap provide pathways into US defense and civil agencies requiring secure, low-latency LLM inference. This public sector expansion represents a significant market opportunity given government AI adoption initiatives. Enterprise customers are increasingly evaluating alternatives to Nvidia-based solutions due to cost and supply constraints, creating opportunities for Groq to capture market share in industries requiring high-performance inference at scale. Geographic expansion Groq secured a $1.5 billion commitment from Saudi Arabia in February 2025 to fund the largest non-hyperscaler inference cluster with over 19,000 LPUs. This positions Groq as the reference platform for Arabic language models and establishes a major presence in the Middle East market. The company launched its first European data center region in Helsinki in July 2025, meeting data sovereignty requirements while leveraging green hydroelectric power for lower operating costs. This European expansion addresses regulatory compliance needs for multinational customers. Groq is exploring partnerships to co-locate LPUs at remote renewable energy sites, allowing the company to monetize stranded energy while adding global capacity cost-effectively. This approach could enable expansion into emerging markets with abundant renewable resources. Risks Nvidia dependency: Groq's non-exclusive technology licensing agreement with Nvidia creates potential conflicts between independent product development and Nvidia's strategic priorities. Following the agreement, Groq's founder Jonathan Ross, President Sunny Madra, and other key team members joined Nvidia to support scaling the licensed technology, while Groq continues operating independently under new CEO Simon Edwards. The non-exclusive nature of the license means Nvidia could potentially use Groq's inference technology to compete more directly with Groq's own GroqCloud and hardware offerings. This arrangement creates strategic tension between Groq's need to differentiate and Nvidia's ability to leverage licensed technology across its broader product portfolio. Software ecosystem: Despite offering OpenAI-compatible APIs, Groq faces the challenge of competing against Nvidia's deeply entrenched CUDA ecosystem. Developers and enterprises have invested heavily in CUDA-based tooling, libraries, and workflows, creating switching costs that superior hardware performance alone may not overcome. The company must continue investing heavily in software development to match the breadth and maturity of established GPU software stacks. Market timing: Groq's success depends on the continued growth of AI inference workloads and the shift from training-focused to inference-focused compute spending. If the AI market experiences a downturn or if new model architectures emerge that favor different hardware approaches, Groq's specialized LPU architecture could become less relevant, potentially stranding the company's significant R&D investments and manufacturing commitments. News Previous Next DISCLAIMERS This report is for information purposes only and is not to be used or considered as an offer or the solicitation of an offer to sell or to buy or subscribe for securities or other financial instruments. Nothing in this report constitutes investment, legal, accounting or tax advice or a representation that any investment or strategy is suitable or appropriate to your individual circumstances or otherwise constitutes a personal trade recommendation to you. This research report has been prepared solely by Sacra and should not be considered a product of any person or entity that makes such report available, if any. Information and opinions presented in the sections of the report were obtained or derived from sources Sacra believes are reliable, but Sacra makes no representation as to their accuracy or completeness. Past performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, express or implied, is made regarding future performance. Information, opinions and estimates contained in this report reflect a determination at its original date of publication by Sacra and are subject to change without notice. Sacra accepts no liability for loss arising from the use of the material presented in this report, except that this exclusion of liability does not apply to the extent that liability arises under specific statutes or regulations applicable to Sacra. Sacra may have issued, and may in the future issue, other reports that are inconsistent with, and reach different conclusions from, the information presented in this report. Those reports reflect different assumptions, views and analytical methods of the analysts who prepared them and Sacra is under no obligation to ensure that such other reports are brought to the attention of any recipient of this report. All rights reserved. All material presented in this report, unless specifically indicated otherwise is under copyright to Sacra. Sacra reserves any and all intellectual property rights in the report. All trademarks, service marks and logos used in this report are trademarks or service marks or registered trademarks or service marks of Sacra. Any modification, copying, displaying, distributing, transmitting, publishing, licensing, creating derivative works from, or selling any report is strictly prohibited. None of the material, nor its content, nor any copy of it, may be altered in any way, transmitted to, copied or distributed to any other party, without the prior express written permission of Sacra. Any unauthorized duplication, redistribution or disclosure of this report will result in prosecution. Filings Name Type Filed At Actions groq_coi_2026-02-10 Certificate of Incorporation Feb 10, 2026 View PDF groq_coi_2025-12-15 Certificate of Incorporation Dec 15, 2025 View PDF groq_coi_2025-12-09 Certificate of Incorporation Dec 09, 2025 View PDF groq_coi_2025-07-25 Certificate of Incorporation Jul 25, 2025 View PDF groq_coi_2024-08-01 Certificate of Incorporation Jan 03, 2024 View PDF Valuation & Funding History Detailed breakdown of all valuation events Date Event Valuation Amount Price Status Jun 2026 Growth Sources Groq Raises 650 Million To Help Startup Pivot After Nvidia Deal bloomberg.com Ai Chipmaker Groq Confirms 650m Raise Re Staffs After Nvidias 20b Not Acqui Hire Deal techcrunch.com — $650.0M — Announced Unlock Groq data Sign up for data access May 2026 Growth — — — Announced Jan 2026 Direct Sale — — — Announced Dec 2025 Direct Sale — — — Announced Dec 2025 Direct Sale — — — Closed Sep 2025 Growth — — — Closed Aug 2024 Series D — — — Closed Apr 2021 Series C — — — Closed Aug 2020 Series B Extension — — — Closed Sep 2018 Series B — — — Closed Apr 2017 Series A — — — Closed Create a free account, or log in. Gain access to limited free reports, models, datasets, interviews and memos. Continue with Google or Email Continue By continuing, you agree to our Privacy Policy and Terms of Service. Free article limit reached. Upgrade your account to get unlimited access to all reports, interviews, and community features. Upgrade Cancel anytime. No commitment required. By continuing, you agree to our Privacy Policy and Terms of Service. Standard membership required. Upgrade your account to get unlimited access to all reports, interviews, and community features Upgrade Cancel anytime. No commitment required. By continuing, you agree to our Privacy Policy and Terms of Service. Standard membership required. Sign up or log in to your Standard account to get unlimited access to all reports, interviews, and community features Continue with Google or Email Continue By continuing, you agree to our Privacy Policy and Terms of Service. View PDF Embed PDF View Model Details Headquarters Mountain View, United States CEO Jonathan Ross Website groq.com Milestones FOUNDING YEAR 2016 Listed In #ai #ai-chips Outline Revenue Valuation & Funding Product Business Model Competition TAM Expansion Risks News Embed Document PDF Embed code is missing! Embed options Height Width By embedding Sacra content on your site, you are agreeing to Sacra Terms of Service Copy The private markets research you need to be a better investor. Privacy Policy | Terms of Service Sacra 2025 © All Rights Reserved Home Features Customers Pricing About Sign Up Login Support Companies Categories Reports Q&A Search Docs Integrations Join the Sacra community and get private markets research and data straight to your inbox. Join",
      "title": "Groq revenue, valuation & funding - Sacra",
      "url": "https://sacra.com/c/groq/"
    },
    {
      "content": "17 Sept 2025 — The company previously raised $640 million at a $2.8 billion valuation in August 2024. This latest funding has more than doubled the AI chip ...",
      "favicon": "https://www.itiger.com/favicon.ico",
      "published_date": "2026-07-27T00:00:00.0000000",
      "score": 0.4849076,
      "text": "17 Sept 2025 — The company previously raised $640 million at a $2.8 billion valuation in August 2024. This latest funding has more than doubled the AI chip ...",
      "title": "LPU Inference Engine Gains Financial Backing! Groq ...",
      "url": "https://www.itiger.com/news/1169807024"
    },
    {
      "content": "30 Jul 2025 — AI chip company Groq is in talks to raise $600 million in a new funding round that would see the company valued at around $6 billion.",
      "favicon": "https://www.datacenterdynamics.com/static/img/favicon-32x32.6fa9e0ecdb87.png",
      "published_date": "2026-07-31T00:00:00.0000000",
      "score": 0.4987998,
      "text": "30 Jul 2025 — AI chip company Groq is in talks to raise $600 million in a new funding round that would see the company valued at around $6 billion.",
      "title": "Groq in talks over fresh funding for $6bn valuation - report",
      "url": "https://www.datacenterdynamics.com/en/news/groq-in-talks-over-fresh-funding-for-6bn-valuation-report/"
    },
    {
      "content": "22 Jun 2026 — San Francisco, CA, June 22, 2026 – Groq today announced $650 million in new growth capital to accelerate the expansion of its AI inference cloud ...",
      "favicon": "https://groq.com/favicon.ico",
      "published_date": "2026-07-22T00:00:00.0000000",
      "score": 0.3905561,
      "text": "22 Jun 2026 — San Francisco, CA, June 22, 2026 – Groq today announced $650 million in new growth capital to accelerate the expansion of its AI inference cloud ...",
      "title": "Groq Raises $650M to Scale Its AI Inference Cloud Business",
      "url": "https://groq.com/newsroom/groq-raises-usd650m-to-scale-its-ai-inference-cloud-business"
    },
    {
      "content": "Language Processing Unit (LPU) Revenue US$500 million (2025) In December 2025, Nvidia and Groq announced an agreement reportedly valued at approximately US$20 ...",
      "favicon": "https://en.wikipedia.org/static/apple-touch/wikipedia.png",
      "published_date": "2026-07-17T00:00:00.0000000",
      "score": 0.4499932,
      "text": "Groq Groq, Inc. is an American artificial intelligence (AI) company that builds an AI accelerator application-specific integrated circuit (ASIC) that they call the Language Processing Unit (LPU) and related hardware to accelerate the inference performance of AI workloads. Examples of the types AI workloads that run on Groq's LPU are: large language models (LLMs), image classification, anomaly detection, and predictive analysis. Groq is headquartered in Mountain View, CA, and has offices in San Jose, CA, Liberty Lake, WA, Toronto, Canada, London, U.K. and remote employees throughout North America and Europe. History. Groq was founded in 2016 by a group of former Google engineers, led by Jonathan Ross, one of the designers of the Tensor Processing Unit (TPU), an AI accelerator ASIC, and Douglas Wightman, an entrepreneur and former engineer at Google X (known as X Development), who served as the company’s first CEO. Groq received seed funding from Social Capital's Chamath Palihapitiya, with a $10 million investment in 2017 and soon after secured additional funding. In April 2021, Groq raised $300 million in a series C round led by Tiger Global Management and D1 Capital Partners. Current investors include: The Spruce House Partnership, Addition, GCM Grosvenor, Xⁿ, Firebolt Ventures, General Global Capital, and Tru Arrow Partners, as well as follow-on investments from TDK Ventures, XTX Ventures, Boardman Bay Capital Management, and Infinitum Partners. After Groq’s series C funding round, it was valued at over $1billion, making the startup a unicorn. On March 1, 2022, Groq acquired Maxeler Technologies, a company known for its dataflow systems technologies. On August 16, 2023, Groq selected Samsung Electronics foundry in Taylor, Texas to manufacture its next generation chips, on Samsung's 4-nanometer (nm) process node. This was the first order at this new Samsung chip factory. On February 19, 2024, Groq soft launched a developer platform, GroqCloud, to attract developers into using the Groq API and rent access to their chips. On March 1, 2024 Groq acquired Definitive Intelligence, a startup known for offering a range of business-oriented AI solutions, to help with its cloud platform. Groq raised $640 million in a series D round led by BlackRock Private Equity Partners in August 2024, valuing the company at $2.8 billion. A 2025 update by Groq on its website reports they have secured 1.5 Billion in funding from the KSA (Kingdom of Saudi Arabia) and a deal with Bell Canada to expand national AI infrastructure. Language Processing Unit. Groq's initial name for their ASIC was the Tensor Streaming Processor (TSP), but later rebranded the TSP as the Language Processing Unit (LPU). The LPU features a functionally sliced microarchitecture, where memory units are interleaved with vector and matrix computation units. This design facilitates the exploitation of dataflow locality in AI compute graphs, improving execution performance and efficiency. The LPU was designed off of two key observations: In addition to its functionally sliced microarchitecture, the LPU can also be characterized by its single core, deterministic architecture. The LPU is able to achieve deterministic execution by avoiding the use of traditional reactive hardware components (branch predictors, arbiters, reordering buffers, caches) and by having all execution explicitly controlled by the compiler thereby guaranteeing determinism in execution of an LPU program. The first generation of the LPU (LPU v1) yields a computational density of more than 1TeraOp/s per square mm of silicon for its 25×29 mm 14nm chip operating at a nominal clock frequency of 900 MHz. The second generation of the LPU (LPU v2) will be manufactured on Samsung's 4nm process node. Performance. Groq emerged as the first API provider to break the 100 tokens per second generation rate while running Meta’s Llama2-70B parameter model. Groq currently hosts a variety of open-source large language models running on its LPUs for public access. Access to these demos are available through Groq's website. The LPU's performance while running these open source LLMs has been independently benchmarked by ArtificialAnalysis.ai, in comparison with other LLM providers. The LPU's measured performance is shown in the table below:",
      "title": "Groq",
      "url": "https://en.wikipedia.org/wiki/Groq"
    }
  ]
}