# Cerebrium **Source:** https://geo.sig.ai/brands/cerebrium **Vertical:** Infrastructure **Subcategory:** Cloud Services **Tier:** Emerging **Website:** cerebrium.ai **Last Updated:** 2026-04-14 ## Summary NY serverless AI inference platform delivering 40% cost savings scaling to 10K+ requests/minute; $9M Google Gradient Ventures seed competing with Modal and Replicate for AI model hosting infrastructure. ## Company Overview Cerebrium is a New York-based serverless cloud infrastructure platform for AI workloads — backed with $9 million raised including an $8.5 million seed round led by Google Gradient Ventures in July 2025 — providing a compute layer where AI companies can deploy, scale, and run AI models (LLMs, speech models, image generation, custom fine-tuned models) at 40% lower cost than AWS and GCP while auto-scaling from zero to 10,000+ requests per minute. Founded in 2021 by Michael Louis and Jonathan Irwin with a lean 4-person engineering team, Cerebrium serves notable AI companies including Tavus, CivitAI, Twilio, and Deepgram with millions in ARR. Cerebrium's serverless inference infrastructure addresses the economics of AI model hosting: running dedicated GPU instances for AI models during low-traffic periods wastes significant compute spend — a model serving 1,000 requests/hour at 3 AM doesn't need the same GPU capacity as the same model at peak hours. Cerebrium's serverless architecture scales model instances to zero during idle periods and spins up additional instances in seconds when demand spikes — providing the economics of pay-per-request without the cold-start latency that makes serverless impractical for latency-sensitive applications. The pre-built model templates (common LLMs, Whisper for speech, Stable Diffusion for image generation) enable sub-5-minute deployment for standard use cases. In 2025, Cerebrium competes in the AI model hosting and serverless inference market with Modal (serverless compute for AI, $45M raised), Replicate (serverless AI model API, $40M raised), and Banana (serverless GPU hosting, $3.1M raised) for AI application inference infrastructure. Google Gradient Ventures' lead on the seed round reflects Google's strategic interest in AI infrastructure that runs on Google Cloud's GPU fleet. The AI inference market has grown explosively as LLM-based applications require scalable model hosting that general-purpose cloud providers (AWS SageMaker, Google Vertex AI) make complex and expensive for lean startup teams. The 2025 strategy focuses on growing the speech and video AI inference vertical (voice cloning, real-time transcription), building the multi-region deployment for latency-sensitive global applications, and expanding the fine-tuned model hosting for enterprises with custom AI models. ## Frequently Asked Questions ### What is Cerebrium? Cerebrium is a serverless cloud infrastructure platform that makes it easy to build and deploy AI applications scalably and performantly. Founded in 2021 and headquartered in New York, the company enables customers to run AI/ML workloads with minimal engineering overhead. ### What products and services does Cerebrium offer? Cerebrium provides serverless AI infrastructure and a cloud platform for AI/ML applications. The platform enables 40% cost savings versus traditional cloud providers while scaling models to 10K+ requests per minute with minimal engineering overhead. ### Who are Cerebrium's customers? Cerebrium serves companies including Tavus, CivitAI, Twilio, and Deepgram. The platform powers multimodal AI workloads such as personalized video content and voice AI applications. ### When was Cerebrium founded and by whom? Cerebrium was founded in 2021 by Michael Louis and Jonathan Irwin, who originated from Cape Town. The company was part of Y Combinator's W22 batch. ### Where is Cerebrium located? Cerebrium is headquartered in New York, New York. The founders originally came from Cape Town before establishing the company in NYC. ### How much funding has Cerebrium raised? Cerebrium has raised $9M in total funding, including an $8.5M seed round in July 2025 led by Gradient Ventures. Other investors include Maxitech, Authentic Ventures, and Y Combinator. ### What are Cerebrium's key achievements and metrics? Cerebrium has achieved millions in annual recurring revenue (ARR) with a lean engineering team of just 4 people. The platform delivers 40% cost savings and can scale models to handle 10K+ requests per minute. ### What is Cerebrium's technology approach? Cerebrium uses a serverless architecture to provide cloud infrastructure for AI and machine learning applications. This approach enables multimodal AI workloads including video and voice processing with minimal engineering overhead. ### What industries does Cerebrium serve? Cerebrium operates in the infrastructure industry, specifically focusing on AI/ML cloud infrastructure. The company serves customers across various sectors including communications, AI content generation, and voice technology. ### What are the latest developments at Cerebrium? As of July 2025, Cerebrium raised an $8.5M seed round led by Gradient Ventures, bringing total funding to $9M. The company has achieved millions in ARR while maintaining a lean team of 4 engineers and serving major customers like Tavus, CivitAI, Twilio, and Deepgram. ## Tags b2b, platform, cloud-native, infrastructure, developer-tools, ai-powered, saas --- *Data from geo.sig.ai Brand Intelligence Database. Updated 2026-04-14.*