Startup perk category
Compare ai infrastructure & compute perks for startups and growing teams.
These listings are also available in the full filterable perks directory.
IBM helps startups at every stage of their journey. With access to IBM’s secure and open hybrid cloud platform (including Red Hat OpenShift), startups are able to build with up to $120,000 in IBM Cloud credits.
RunPod for Startups gives venture-backed AI companies dedicated GPU compute credits and infrastructure support - the high-performance cloud GPU platform for model training; fine-tuning; and production inference at scale.
Alibaba Cloud gives qualifying early-stage startups up to $50000 in cloud credits; technical support; co-marketing opportunities; and access to Alibaba global investor and partner networks.
Together AI Startup Accelerator gives qualifying AI-native companies free compute credits; dedicated engineering support; and GTM amplification to 800K+ developers on the leading open-source AI cloud.
NVIDIA Inception is a free startup program giving AI and data science companies preferred pricing on NVIDIA hardware and software; co-marketing support; and connections to NVIDIA experts; VCs; and the global Inception ecosystem.
Cerebras for Startups gives qualifying AI-focused early-stage companies up to $22500 in Cerebras compute credits; priority access to new AI models; and technical and marketing support — the fastest AI inference platform powered by the Wafer Scale Engine.
AWS Activate Portfolio+ gives startups in select top-tier AWS partner organisations up to $100000 in AWS credits — the highest tier of AWS Activate for high-growth startups.
If you’re looking for a new way of working and want more than the status quo, Google Cloud is the new way to cloud. With the AI, data, infrastructure, collaboration, and security tools built for today and tomorrow. AI that’s ready for what’s next, data insights that help innovate faster, infrastructure that’s designed to meet industry needs, collaboration tools that help teams do their best work, and security that proactively stops threats. Discover what’s new with Google Cloud.
Databricks for Startups gives qualifying VC-funded early-stage companies up to $50000 in Databricks credits; free business-tier technical support; and co-marketing opportunities on the unified Data and AI platform.
If you’re looking for a new way of working and want more than the status quo, Google Cloud is the new way to cloud. With the AI, data, infrastructure, collaboration, and security tools built for today and tomorrow. AI that’s ready for what’s next, data insights that help innovate faster, infrastructure that’s designed to meet industry needs, collaboration tools that help teams do their best work, and security that proactively stops threats. Discover what’s new with Google Cloud.
If you’re looking for a new way of working and want more than the status quo, Google Cloud is the new way to cloud. With the AI, data, infrastructure, collaboration, and security tools built for today and tomorrow. AI that’s ready for what’s next, data insights that help innovate faster, infrastructure that’s designed to meet industry needs, collaboration tools that help teams do their best work, and security that proactively stops threats. Discover what’s new with Google Cloud.
Render's Scale AI tier gives AI and compute-intensive startups with over $2.5M raised up to $100,000 in cloud hosting credits — one of the highest non-hyperscaler credit amounts available to early-stage AI companies.
Roboflow for Startups gives qualifying early-stage companies free access to the Roboflow computer vision platform - the end-to-end tool for building; training; and deploying vision AI models with no infrastructure overhead.
Roboflow for Startups gives qualifying early-stage companies free access to the Roboflow computer vision platform - the end-to-end tool for building; training; and deploying vision AI models with no infrastructure overhead.
Databricks for Startups gives qualifying VC-funded early-stage companies up to $50000 in Databricks credits; free business-tier technical support; and co-marketing opportunities on the unified Data and AI platform.
Nebius for Startups gives AI-focused early-stage companies $5,000 in immediate GPU credits plus up to $100,000 in discounted compute — enterprise-grade H100 infrastructure without hyperscaler friction or quota roadblocks.
Nebius for Startups gives AI-focused early-stage companies $5,000 in immediate GPU credits plus up to $100,000 in discounted compute — enterprise-grade H100 infrastructure without hyperscaler friction or quota roadblocks.
OpenRouter for Startups gives qualifying early-stage companies $5000 in universal inference credits — usable across 300+ frontier and open-source AI models with 0% processing fees; through a single unified API.
Groq for Startups gives qualifying early-stage AI companies $10000 in free inference compute credits — the world's fastest AI inference platform powered by the Language Processing Unit (LPU) delivering up to 2000 tokens per second.
DeepInfra DeepStart gives qualifying early-stage AI companies up to 1 billion free inference tokens — the cost-effective ML inference platform for running state-of-the-art open-source models including Llama; Mistral; Qwen; and Stable Diffusion in production.
Modal for Startups gives qualifying early-stage AI companies $25000 in compute credits — the serverless cloud platform for deploying custom AI models and Python workloads with GPU acceleration; auto-scaling; and instant cold-start.
DigitalOcean Hatch's AI/ML tier gives AI-focused startups up to $100,000 in compute credits plus preferred GPU Droplet pricing at $1.90/hr — making it one of the most accessible GPU compute programs for AI startups. Contact hatch@digitalocean.com for AI tier details.
Together AI's startup accelerator gives startups that have raised between $5M and $10M up to $30,000 in compute credits plus 6 hours of engineering support — covering serverless inference, fine-tuning, and custom kernel optimisation on the largest open-weight model fleet.
Together AI's startup accelerator gives startups that have raised over $10M up to $50,000 in compute credits plus 10 hours of engineering support — covering serverless inference, fine-tuning, and custom kernel optimisation on the largest open-weight model fleet.