Inference hosting
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | AI territory |
|---|---|
| Categories | 6 |
| Companies | 31 |
| Map changed | 2026-11-19 |
| Map | open the map |
Inference hosting is a territory of the AI industry on map of ai. In short: inference hosting provides api access to run foundational large language models on remote servers, returning generated text or embeddings. The map lists 31 products and companies in 6 categories.
The map of inference hosting has 6 categories. Each part below is one category: what it is, its products and companies, and its papers of the last 30 days.
serverless gpu inference6 products and companies · its own article →
Serverless GPU inference provides on-demand GPU compute for running machine learning models without managing infrastructure; used by developers, startups, and enterprises running inference workloads that don't justify dedicated hardware.
Open ground
Replicate, Together AI, Banana, and Hugging Face all offer pay-per-inference or hourly GPU billing; none of them offers reserved GPU capacity with committed spend discounts comparable to hyperscaler reserved instances.
open model api providers7 products and companies
Open model API providers host inference endpoints for large language models and other AI models, allowing developers and applications to query models via REST or other APIs without managing infrastructure. Used by startups, enterprises, and developers building AI features who want managed model access.
Products and companies
custom silicon inference3 products and companies
Custom silicon inference refers to dedicated hardware processors (ASICs or specialized chips) designed and optimized for running machine learning inference workloads. Data centers, cloud providers, and enterprises deploy these chips to reduce latency and energy consumption when running inference at scale.
hyperscaler model platforms5 products and companies
Platforms that host and serve large language models and other generative AI models for inference at scale, used by developers and enterprises to run AI applications without managing infrastructure. These platforms provide APIs, SDKs, and dashboards to send prompts and receive model outputs.
Products and companies: amazon bedrock, azure ai foundry, google vertex ai, oracle oci generative ai, ibm watsonx.
inference engines5 products and companies
Inference hosting provides compute infrastructure where organizations run machine learning models to generate predictions or completions on demand. Machine learning teams, application developers, and enterprises use these services to execute trained models without managing hardware.
Products and companies: vllm, sglang, ollama, llama.cpp, nvidia nim.
Papers, last 30 days:model routers and gateways5 products and companies · its own article →
Model routers and gateways direct inference requests across multiple LLM endpoints and providers, selecting targets based on cost, latency, or model capability. ML teams, application developers, and enterprises use them to avoid vendor lock-in and optimize inference spend.
Products and companies: openrouter, litellm, portkey, cloudflare workers ai, vercel ai gateway.
See also
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong.