FriendliAI runs its inference stack and APIs on Nebius infrastructure. Its November 2025 announcement describes integrating its proprietary inference optimization technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity.
FriendliAI says the combined offering automatically scales GPU resources with inference demand and guarantees GPU availability for sudden spikes or large workloads. The announcement advertises over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee. These are claims about the combined offering. Its October announcement presents customer support bots, coding assistants and AI agents as applications served through FriendliAI APIs on Nebius.
FriendliAI Partners with Nebius to Deliver High-Performance, Cost-Efficient AI Inference
Published 19 Nov 2025 · FriendliAI · Web page
FriendliAI describes integrating its inference technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity. It advertises automatic GPU scaling, guaranteed availability, over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee for the combined offering.
FriendliAI Partners with Nebius to Deliver High-Performance, Cost-Efficient AI InferencePublished 19 Nov 2025 · FriendliAI · Web pageFriendliAI describes integrating its inference technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity. It advertises automatic GPU scaling, guaranteed availability, over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee for the combined offering.
FriendliAI Expands Ultra-Fast AI Inference Platform with Nebius AI Cloud Integration
Listed 21 Oct 2026 · The National Law Review · Web page
FriendliAI’s syndicated press release describes its APIs running on Nebius infrastructure for inference. Its title names Nebius AI Cloud Integration. The publication header states October 28, 2025.
FriendliAI Expands Ultra-Fast AI Inference Platform with Nebius AI Cloud IntegrationListed 21 Oct 2026 · The National Law Review · Web pageFriendliAI’s syndicated press release describes its APIs running on Nebius infrastructure for inference. Its title names Nebius AI Cloud Integration. The publication header states October 28, 2025.
FriendliAI announces its Nebius collaboration
Published 28 Oct 2025 · LinkedIn · Web page
FriendliAI announces combining its inference platform with Nebius GPU Cloud. It presents bots, coding assistants, agentic systems and simulations as target applications without identifying downstream deployments.
FriendliAI announces its Nebius collaborationPublished 28 Oct 2025 · LinkedIn · Web pageFriendliAI announces combining its inference platform with Nebius GPU Cloud. It presents bots, coding assistants, agentic systems and simulations as target applications without identifying downstream deployments.
FriendliAI founder comments on Nebius partnership
Published 28 Oct 2025 · LinkedIn · Web page
Byung-Gon Chun announces FriendliAI’s partnership with Nebius to scale AI workloads and shares FriendliAI’s company announcement.
FriendliAI founder comments on Nebius partnershipPublished 28 Oct 2025 · LinkedIn · Web pageByung-Gon Chun announces FriendliAI’s partnership with Nebius to scale AI workloads and shares FriendliAI’s company announcement.
FriendliAI, operated by FriendliAI Corp., provides infrastructure for deploying and running open-weight and custom AI models. Its customers include developers, startups and enterprises building coding assistants, AI agents, chatbots, search and multimodal applications. Its main products are Model APIs for managed inference, Dedicated Endpoints for dedicated model serving, and BYOG, which runs its inference stack on customers’ own cloud GPUs. The platform handles deployment, scaling and performance optimization, with customers including LG AI Research, SK Telecom and Upstage.
FriendliAI adds NVIDIA Nemotron Retriever models to Dedicated Endpoints
FriendliAI made NVIDIA Nemotron 3.5 Embed 1B, Embed 8B and Rerank 8B available on Dedicated Endpoints. The models support retrieval across text and visual content for enterprise search and AI agents.
FriendliAI makes NVIDIA Nemotron 3.5 Super VL available
FriendliAI added NVIDIA Nemotron 3.5 Super VL to Dedicated Endpoints. The multimodal model combines visual understanding, reasoning and tool use for workloads involving text, images, documents and video.
FriendliAI launches FriendliLink for coding agents
FriendliAI introduced FriendliLink, a command-line tool that connects coding agents to Friendli Model APIs. It supports model discovery, configuration changes and restoration of previous settings across agents including Claude Code, Codex and Cursor.
FriendliAI provides AI model deployment, inference APIs and dedicated endpoints. · Infrastructure
How they use Nebius AI Cloud
FriendliAI runs its inference stack and APIs on Nebius infrastructure. Its November 2025 announcement describes integrating its proprietary inference optimization technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity.
FriendliAI says the combined offering automatically scales GPU resources with inference demand and guarantees GPU availability for sudden spikes or large workloads. The announcement advertises over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee. These are claims about the combined offering. Its October announcement presents customer support bots, coding assistants and AI agents as applications served through FriendliAI APIs on Nebius.
FriendliAI Partners with Nebius to Deliver High-Performance, Cost-Efficient AI Inference
Published 19 Nov 2025 · FriendliAI · Web page
FriendliAI describes integrating its inference technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity. It advertises automatic GPU scaling, guaranteed availability, over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee for the combined offering.
FriendliAI Partners with Nebius to Deliver High-Performance, Cost-Efficient AI InferencePublished 19 Nov 2025 · FriendliAI · Web pageFriendliAI describes integrating its inference technology with Nebius GPU infrastructure and adding Nebius to its multicloud capacity. It advertises automatic GPU scaling, guaranteed availability, over 50% lower GPU costs, over twice the inference speed and a 99.99% uptime guarantee for the combined offering.
FriendliAI Expands Ultra-Fast AI Inference Platform with Nebius AI Cloud Integration
Listed 21 Oct 2026 · The National Law Review · Web page
FriendliAI’s syndicated press release describes its APIs running on Nebius infrastructure for inference. Its title names Nebius AI Cloud Integration. The publication header states October 28, 2025.
FriendliAI Expands Ultra-Fast AI Inference Platform with Nebius AI Cloud IntegrationListed 21 Oct 2026 · The National Law Review · Web pageFriendliAI’s syndicated press release describes its APIs running on Nebius infrastructure for inference. Its title names Nebius AI Cloud Integration. The publication header states October 28, 2025.
FriendliAI announces its Nebius collaboration
Published 28 Oct 2025 · LinkedIn · Web page
FriendliAI announces combining its inference platform with Nebius GPU Cloud. It presents bots, coding assistants, agentic systems and simulations as target applications without identifying downstream deployments.
FriendliAI announces its Nebius collaborationPublished 28 Oct 2025 · LinkedIn · Web pageFriendliAI announces combining its inference platform with Nebius GPU Cloud. It presents bots, coding assistants, agentic systems and simulations as target applications without identifying downstream deployments.
FriendliAI founder comments on Nebius partnership
Published 28 Oct 2025 · LinkedIn · Web page
Byung-Gon Chun announces FriendliAI’s partnership with Nebius to scale AI workloads and shares FriendliAI’s company announcement.
FriendliAI founder comments on Nebius partnershipPublished 28 Oct 2025 · LinkedIn · Web pageByung-Gon Chun announces FriendliAI’s partnership with Nebius to scale AI workloads and shares FriendliAI’s company announcement.
FriendliAI, operated by FriendliAI Corp., provides infrastructure for deploying and running open-weight and custom AI models. Its customers include developers, startups and enterprises building coding assistants, AI agents, chatbots, search and multimodal applications. Its main products are Model APIs for managed inference, Dedicated Endpoints for dedicated model serving, and BYOG, which runs its inference stack on customers’ own cloud GPUs. The platform handles deployment, scaling and performance optimization, with customers including LG AI Research, SK Telecom and Upstage.
FriendliAI adds NVIDIA Nemotron Retriever models to Dedicated Endpoints
FriendliAI made NVIDIA Nemotron 3.5 Embed 1B, Embed 8B and Rerank 8B available on Dedicated Endpoints. The models support retrieval across text and visual content for enterprise search and AI agents.
FriendliAI makes NVIDIA Nemotron 3.5 Super VL available
FriendliAI added NVIDIA Nemotron 3.5 Super VL to Dedicated Endpoints. The multimodal model combines visual understanding, reasoning and tool use for workloads involving text, images, documents and video.
FriendliAI launches FriendliLink for coding agents
FriendliAI introduced FriendliLink, a command-line tool that connects coding agents to Friendli Model APIs. It supports model discovery, configuration changes and restoration of previous settings across agents including Claude Code, Codex and Cursor.
