what's Hyperbolic Labs?
Hyperbolic Labs is building an open-access AI cloud designed to democratize access to AI resources by aggregating global GPU compute power. The platform addresses key challenges in AI deployment, including scalability through a distributed system that optimizes GPU performance, verification to ensure outputs match specified AI models, and privacy protections for user data processed across third-party nodes. By leveraging blockchain technology, Hyperbolic enables a decentralized economy where contributors are rewarded, making high-performance AI tools available to developers without reliance on centralized providers. The project focuses on creating an ecosystem for AI services, including inference and training on models like DeepSeek-V3. It supports a range of GPUs, from high-end H100s and H200s to consumer-grade options like RTX 4090s, with instant rental capabilities and competitive pricing. Hyperbolic's verifiable AI infrastructure ensures trust in computations, positioning it as a bridge between blockchain and AI to foster innovation in decentralized applications. Through strategic funding and partnerships, Hyperbolic aims to scale supply and reliability, offering features like referral programs and educational credits to broaden adoption. Its commitment to open-source models and rapid deployment underscores a mission to make AI accessible and equitable for builders worldwide.
Hyperbolic Labs partners with Ornn Exchange for regulated compute derivatives using real-time GPU data.
Partners with Ornn Exchange to power a regulated compute derivatives market using real-time GPU infrastructure data.
Powers the compute infrastructure behind TTT-Discover, an open model system using test-time learning for scientific discovery problems.
Project pivoted to Web2 and is described as DOA (dead on arrival).
Platform now supports organization setup functionality for team management.
DeepSeek-Math-V2 model is now running on 8xH200 node on the on-demand GPU platform.
Announces availability of NVIDIA's Nemotron Models on the platform, offering open models and techniques for specialized agentic AI applications.
Listed as inference provider on Hugging Face, offering Omni service.
Project is listed among AI projects planning Token Generation Event in Q4, providing more specific timing than previously indicated.
Third-party analysis shows the gpt-oss-120b model achieves 402 tokens/sec with consistent performance across context lengths, positioning alongside top inference providers.
Launches Hyper-dOS platform with auto-scaling GPU clusters and powers Klok app GPU marketplace.
TGE planned for coming months, listed among top-tier infrastructure projects.
New referral program launches offering $5/$6 rewards for signups, alongside H100 GPU rentals from $0.99/hr.
Project receives backing from Polychain Capital, known for organic growth approach and 20.08X portfolio ROI.
GPU Marketplace adds Custom Port Configuration functionality
AI-focused protocol with $19M in funding requires $5 minimum deposit for participation in GPT interactions and mission-based rewards system.
Points farming available through interaction with AI models.
Selected for IOSG Kickstarter Program alongside major crypto projects including Aztec Network, CoinGecko, and Mantle.
Providing cost-efficient inference for Deepseek AI models, featured on Lex Fridman podcast alongside Together AI and Fireworks AI.
AI project focused on bridging Web2 and Web3, emphasizing high-performance AI and reliable infrastructure.