# Aguasto

> Aguasto is an emerging, U.S.-focused AI compute company. AI Compute. Built From the Ground Up.

The development strategy connects modular high-density compute facilities, planned dedicated GPU capacity, and planned enterprise inference for supported open-weight models. Independently operated facilities would require utility-authorized service, local approvals and integrated commissioning. GPU capacity is not currently bookable. No public inference API, model endpoint, or token pricing is available. Facility and network diagrams are illustrative concepts, not operating assets or actual project locations.

## Site discovery

- [Canonical website](https://www.aguasto.com/)
- [Full public reference](https://www.aguasto.com/llms-full.txt): Build-derived content from the published informational pages in one document. Interactive tools and inquiry instructions remain at their canonical HTML links.
- [XML sitemap](https://www.aguasto.com/sitemap.xml): 89 published canonical HTML pages. Markdown representations are alternate documents and are not listed as separate search pages.
- [Crawler policy](https://www.aguasto.com/robots.txt): Search and retrieval access is distinct from model-training access. Read the actual policy before crawling.
- [RSS publications](https://www.aguasto.com/feed.xml) and [Atom publications](https://www.aguasto.com/atom.xml): published editorial records only, with original publication dates. Drafts and product concepts are excluded.

## Public website

- [API versioning policy](https://www.aguasto.com/developers/versioning)
- [Developer sandbox](https://www.aguasto.com/sandbox)
- [Home](https://www.aguasto.com/)
- [Infrastructure](https://www.aguasto.com/infrastructure)
- [GPU Cloud](https://www.aguasto.com/gpu-cloud)
- [AI Inference](https://www.aguasto.com/ai-inference)
- [Development & Investment](https://www.aguasto.com/development)
- [Company](https://www.aguasto.com/company)
- [Contact](https://www.aguasto.com/contact)
- [Privacy](https://www.aguasto.com/privacy)
- [Research topics](https://www.aguasto.com/topics)
- [GPU systems & local clusters](https://www.aguasto.com/topics/gpu-systems)
- [Model serving & inference](https://www.aguasto.com/topics/model-serving)
- [Modular infrastructure & resources](https://www.aguasto.com/topics/modular-infrastructure)
- [Compute economics & evidence](https://www.aguasto.com/topics/compute-economics)
- [Glossary](https://www.aguasto.com/glossary)
- [Methodologies](https://www.aguasto.com/methodologies)
- [Developers](https://www.aguasto.com/developers)
- [Resources](https://www.aguasto.com/resources)
- [GPU Library](https://www.aguasto.com/gpus)
- [Model Index](https://www.aguasto.com/models)
- [Inference Stack](https://www.aguasto.com/inference-stack)
- [Learn](https://www.aguasto.com/learn)
- [Insights](https://www.aguasto.com/insights)
- [Research](https://www.aguasto.com/research)
- [Industry cases](https://www.aguasto.com/research/cases)
- [Why Modular](https://www.aguasto.com/infrastructure/why-modular)
- [Cooling & Resource Efficiency](https://www.aguasto.com/infrastructure/cooling-resource-efficiency)
- [Partners](https://www.aguasto.com/partners)
- [GPU comparison](https://www.aguasto.com/gpus/compare)
- [Model comparison](https://www.aguasto.com/models/compare)
- [Engineering tools](https://www.aguasto.com/tools)
- [Development friction](https://www.aguasto.com/research/hyperscale-development-friction)
- [NVIDIA H100 SXM](https://www.aguasto.com/gpus/h100)
- [NVIDIA H200 SXM](https://www.aguasto.com/gpus/h200)
- [NVIDIA B200 SXM](https://www.aguasto.com/gpus/b200)
- [NVIDIA B300](https://www.aguasto.com/gpus/b300)
- [NVIDIA GB200 NVL72](https://www.aguasto.com/gpus/gb200-nvl72)
- [NVIDIA GB300 NVL72](https://www.aguasto.com/gpus/gb300-nvl72)
- [NVIDIA Vera Rubin NVL72](https://www.aguasto.com/gpus/vera-rubin-nvl72)
- [AMD Instinct MI300X](https://www.aguasto.com/gpus/mi300x)
- [AMD Instinct MI325X](https://www.aguasto.com/gpus/mi325x)
- [AMD Instinct MI355X](https://www.aguasto.com/gpus/mi355x)
- [AMD Instinct MI455X](https://www.aguasto.com/gpus/mi455x)
- [AMD Helios](https://www.aguasto.com/gpus/helios)
- [GLM-4.5](https://www.aguasto.com/models/glm-4-5)
- [Qwen3-235B-A22B](https://www.aguasto.com/models/qwen3-235b-a22b)
- [DeepSeek-R1](https://www.aguasto.com/models/deepseek-r1)
- [Llama 3.3 70B Instruct](https://www.aguasto.com/models/llama-3-3-70b-instruct)
- [Gemma 3 27B IT](https://www.aguasto.com/models/gemma-3-27b-it)
- [gpt-oss-120b](https://www.aguasto.com/models/gpt-oss-120b)
- [Kimi-K2-Instruct](https://www.aguasto.com/models/kimi-k2-instruct)
- [GLM-5.3](https://www.aguasto.com/models/glm-5-3)
- [Qwen3.8-27B](https://www.aguasto.com/models/qwen3-8-27b)
- [DeepSeek-V4.1-Flash](https://www.aguasto.com/models/deepseek-v4-1-flash)
- [Llama 4 Scout 17B-16E Instruct](https://www.aguasto.com/models/llama-4-scout-17b-16e-instruct)
- [Gemma 4 31B IT](https://www.aguasto.com/models/gemma-4-31b-it)
- [Kimi K3](https://www.aguasto.com/models/kimi-k3)
- [vLLM](https://www.aguasto.com/inference-stack/vllm)
- [SGLang](https://www.aguasto.com/inference-stack/sglang)
- [TensorRT-LLM](https://www.aguasto.com/inference-stack/tensorrt-llm)
- [NVIDIA Dynamo](https://www.aguasto.com/inference-stack/dynamo)
- [NVIDIA Triton Inference Server](https://www.aguasto.com/inference-stack/triton)
- [AMD ROCm](https://www.aguasto.com/inference-stack/rocm)
- [Ray Serve](https://www.aguasto.com/inference-stack/ray-serve)
- [Kubernetes GPU orchestration — NVIDIA device plugin](https://www.aguasto.com/inference-stack/kubernetes)
- [Model weights are only the beginning of the memory budget](https://www.aguasto.com/learn/model-weights-and-runtime-memory)
- [Why tokens per second is not enough to benchmark inference](https://www.aguasto.com/research/inference-benchmark-methodology)
- [Prefill, decode and the cost of separating them](https://www.aguasto.com/learn/prefill-decode-and-disaggregated-serving)
- [Choosing between vLLM, SGLang and TensorRT-LLM](https://www.aguasto.com/learn/choosing-an-inference-serving-framework)
- [From Power to Inference—Building a Distributed AI Compute Network](https://www.aguasto.com/learn/from-power-to-inference)
- [Dedicated compute and shared inference: compare the whole workload](https://www.aguasto.com/insights/dedicated-and-shared-inference-economics)
- [Distributed inference reliability begins with the failure boundary](https://www.aguasto.com/learn/distributed-inference-reliability)
- [Core Scientific: reading contracted, energized and billed capacity](https://www.aguasto.com/research/cases/core-scientific-capacity-disclosures)
- [Why Modular AI Infrastructure Matters Now](https://www.aguasto.com/research/why-modular-ai-infrastructure-matters-now)
- [What Is a Modular AI Data Center?](https://www.aguasto.com/learn/what-is-a-modular-ai-data-center)
- [How Modular AI Data Centers Are Built and Commissioned](https://www.aguasto.com/learn/modular-data-center-factory-to-commissioning)
- [Liquid Cooling for High-Density GPU Systems: Follow the Interfaces](https://www.aguasto.com/learn/liquid-cooling-gpu-rack-interfaces)
- [Why Electricity Availability Shapes AI Infrastructure Deployment](https://www.aguasto.com/research/electricity-availability-and-time-to-power)
- [What Is KV Cache? Capacity, Reuse and Isolation](https://www.aguasto.com/learn/kv-cache-capacity-and-policy)
- [Continuous Batching: Throughput Without Hiding Latency](https://www.aguasto.com/learn/continuous-batching-and-latency)
- [Applied Digital: Reading the Polaris Forge 1 Delivery Milestones](https://www.aguasto.com/research/cases/applied-digital-polaris-forge-capacity-milestones)
- [Crusoe Abilene: Separate Delivered Infrastructure from Expansion Plans](https://www.aguasto.com/research/cases/crusoe-abilene-delivery-and-expansion)
- [Data Center Cooling: Choosing Air, Liquid, Dry and Hybrid Architectures](https://www.aguasto.com/learn/data-center-cooling-air-liquid-dry-hybrid)
- [Can Data Centers Reduce Water Consumption? Define the Water Boundary First](https://www.aguasto.com/research/data-center-water-consumption-and-withdrawal)
- [Measuring PUE, WUE and AI Compute Efficiency at the Right Boundary](https://www.aguasto.com/learn/measuring-pue-wue-and-ai-compute-efficiency)
- [The Environmental Trade-offs of High-Density AI Infrastructure](https://www.aguasto.com/research/environmental-tradeoffs-and-infrastructure-reuse)
- [Selecting GPU Systems for Enterprise Inference](https://www.aguasto.com/learn/selecting-enterprise-inference-gpu-systems)
- [Networking Rack-Scale AI Systems: Define the Communication Boundary](https://www.aguasto.com/learn/rack-scale-ai-networking-boundaries)

## Focused section indexes

- [Developers and agent access](https://www.aguasto.com/developers/llms.txt)
- [GPU Cloud](https://www.aguasto.com/gpu-cloud/llms.txt)
- [AI Inference](https://www.aguasto.com/ai-inference/llms.txt)
- [Modular infrastructure](https://www.aguasto.com/infrastructure/llms.txt)
- [Research and resource navigation](https://www.aguasto.com/resources/llms.txt)
- [GPU Library](https://www.aguasto.com/gpus/llms.txt)
- [Model Index](https://www.aguasto.com/models/llms.txt)
- [Inference Stack](https://www.aguasto.com/inference-stack/llms.txt)
- [Research and public industry cases](https://www.aguasto.com/research/llms.txt)
- [Engineering guides](https://www.aguasto.com/learn/llms.txt)
- [Topic navigation](https://www.aguasto.com/topics/llms.txt)

## Markdown representations

These documents are generated from the same built, public HTML content. The linked HTML pages remain canonical.

- [API versioning policy](https://www.aguasto.com/developers/versioning.md)
- [Developer sandbox](https://www.aguasto.com/sandbox.md)
- [Home](https://www.aguasto.com/index.md)
- [Infrastructure](https://www.aguasto.com/infrastructure.md)
- [GPU Cloud](https://www.aguasto.com/gpu-cloud.md)
- [AI Inference](https://www.aguasto.com/ai-inference.md)
- [Development & Investment](https://www.aguasto.com/development.md)
- [Company](https://www.aguasto.com/company.md)
- [Research topics](https://www.aguasto.com/topics.md)
- [GPU systems & local clusters](https://www.aguasto.com/topics/gpu-systems.md)
- [Model serving & inference](https://www.aguasto.com/topics/model-serving.md)
- [Modular infrastructure & resources](https://www.aguasto.com/topics/modular-infrastructure.md)
- [Compute economics & evidence](https://www.aguasto.com/topics/compute-economics.md)
- [Glossary](https://www.aguasto.com/glossary.md)
- [Methodologies](https://www.aguasto.com/methodologies.md)
- [Developers](https://www.aguasto.com/developers.md)
- [Resources](https://www.aguasto.com/resources.md)
- [GPU Library](https://www.aguasto.com/gpus.md)
- [Model Index](https://www.aguasto.com/models.md)
- [Inference Stack](https://www.aguasto.com/inference-stack.md)
- [Learn](https://www.aguasto.com/learn.md)
- [Insights](https://www.aguasto.com/insights.md)
- [Research](https://www.aguasto.com/research.md)
- [Industry cases](https://www.aguasto.com/research/cases.md)
- [Why Modular](https://www.aguasto.com/infrastructure/why-modular.md)
- [Cooling & Resource Efficiency](https://www.aguasto.com/infrastructure/cooling-resource-efficiency.md)
- [Partners](https://www.aguasto.com/partners.md)
- [Development friction](https://www.aguasto.com/research/hyperscale-development-friction.md)
- [NVIDIA H100 SXM](https://www.aguasto.com/gpus/h100.md)
- [NVIDIA H200 SXM](https://www.aguasto.com/gpus/h200.md)
- [NVIDIA B200 SXM](https://www.aguasto.com/gpus/b200.md)
- [NVIDIA B300](https://www.aguasto.com/gpus/b300.md)
- [NVIDIA GB200 NVL72](https://www.aguasto.com/gpus/gb200-nvl72.md)
- [NVIDIA GB300 NVL72](https://www.aguasto.com/gpus/gb300-nvl72.md)
- [NVIDIA Vera Rubin NVL72](https://www.aguasto.com/gpus/vera-rubin-nvl72.md)
- [AMD Instinct MI300X](https://www.aguasto.com/gpus/mi300x.md)
- [AMD Instinct MI325X](https://www.aguasto.com/gpus/mi325x.md)
- [AMD Instinct MI355X](https://www.aguasto.com/gpus/mi355x.md)
- [AMD Instinct MI455X](https://www.aguasto.com/gpus/mi455x.md)
- [AMD Helios](https://www.aguasto.com/gpus/helios.md)
- [GLM-4.5](https://www.aguasto.com/models/glm-4-5.md)
- [Qwen3-235B-A22B](https://www.aguasto.com/models/qwen3-235b-a22b.md)
- [DeepSeek-R1](https://www.aguasto.com/models/deepseek-r1.md)
- [Llama 3.3 70B Instruct](https://www.aguasto.com/models/llama-3-3-70b-instruct.md)
- [Gemma 3 27B IT](https://www.aguasto.com/models/gemma-3-27b-it.md)
- [gpt-oss-120b](https://www.aguasto.com/models/gpt-oss-120b.md)
- [Kimi-K2-Instruct](https://www.aguasto.com/models/kimi-k2-instruct.md)
- [GLM-5.3](https://www.aguasto.com/models/glm-5-3.md)
- [Qwen3.8-27B](https://www.aguasto.com/models/qwen3-8-27b.md)
- [DeepSeek-V4.1-Flash](https://www.aguasto.com/models/deepseek-v4-1-flash.md)
- [Llama 4 Scout 17B-16E Instruct](https://www.aguasto.com/models/llama-4-scout-17b-16e-instruct.md)
- [Gemma 4 31B IT](https://www.aguasto.com/models/gemma-4-31b-it.md)
- [Kimi K3](https://www.aguasto.com/models/kimi-k3.md)
- [vLLM](https://www.aguasto.com/inference-stack/vllm.md)
- [SGLang](https://www.aguasto.com/inference-stack/sglang.md)
- [TensorRT-LLM](https://www.aguasto.com/inference-stack/tensorrt-llm.md)
- [NVIDIA Dynamo](https://www.aguasto.com/inference-stack/dynamo.md)
- [NVIDIA Triton Inference Server](https://www.aguasto.com/inference-stack/triton.md)
- [AMD ROCm](https://www.aguasto.com/inference-stack/rocm.md)
- [Ray Serve](https://www.aguasto.com/inference-stack/ray-serve.md)
- [Kubernetes GPU orchestration — NVIDIA device plugin](https://www.aguasto.com/inference-stack/kubernetes.md)
- [Model weights are only the beginning of the memory budget](https://www.aguasto.com/learn/model-weights-and-runtime-memory.md)
- [Why tokens per second is not enough to benchmark inference](https://www.aguasto.com/research/inference-benchmark-methodology.md)
- [Prefill, decode and the cost of separating them](https://www.aguasto.com/learn/prefill-decode-and-disaggregated-serving.md)
- [Choosing between vLLM, SGLang and TensorRT-LLM](https://www.aguasto.com/learn/choosing-an-inference-serving-framework.md)
- [From Power to Inference—Building a Distributed AI Compute Network](https://www.aguasto.com/learn/from-power-to-inference.md)
- [Dedicated compute and shared inference: compare the whole workload](https://www.aguasto.com/insights/dedicated-and-shared-inference-economics.md)
- [Distributed inference reliability begins with the failure boundary](https://www.aguasto.com/learn/distributed-inference-reliability.md)
- [Core Scientific: reading contracted, energized and billed capacity](https://www.aguasto.com/research/cases/core-scientific-capacity-disclosures.md)
- [Why Modular AI Infrastructure Matters Now](https://www.aguasto.com/research/why-modular-ai-infrastructure-matters-now.md)
- [What Is a Modular AI Data Center?](https://www.aguasto.com/learn/what-is-a-modular-ai-data-center.md)
- [How Modular AI Data Centers Are Built and Commissioned](https://www.aguasto.com/learn/modular-data-center-factory-to-commissioning.md)
- [Liquid Cooling for High-Density GPU Systems: Follow the Interfaces](https://www.aguasto.com/learn/liquid-cooling-gpu-rack-interfaces.md)
- [Why Electricity Availability Shapes AI Infrastructure Deployment](https://www.aguasto.com/research/electricity-availability-and-time-to-power.md)
- [What Is KV Cache? Capacity, Reuse and Isolation](https://www.aguasto.com/learn/kv-cache-capacity-and-policy.md)
- [Continuous Batching: Throughput Without Hiding Latency](https://www.aguasto.com/learn/continuous-batching-and-latency.md)
- [Applied Digital: Reading the Polaris Forge 1 Delivery Milestones](https://www.aguasto.com/research/cases/applied-digital-polaris-forge-capacity-milestones.md)
- [Crusoe Abilene: Separate Delivered Infrastructure from Expansion Plans](https://www.aguasto.com/research/cases/crusoe-abilene-delivery-and-expansion.md)
- [Data Center Cooling: Choosing Air, Liquid, Dry and Hybrid Architectures](https://www.aguasto.com/learn/data-center-cooling-air-liquid-dry-hybrid.md)
- [Can Data Centers Reduce Water Consumption? Define the Water Boundary First](https://www.aguasto.com/research/data-center-water-consumption-and-withdrawal.md)
- [Measuring PUE, WUE and AI Compute Efficiency at the Right Boundary](https://www.aguasto.com/learn/measuring-pue-wue-and-ai-compute-efficiency.md)
- [The Environmental Trade-offs of High-Density AI Infrastructure](https://www.aguasto.com/research/environmental-tradeoffs-and-infrastructure-reuse.md)
- [Selecting GPU Systems for Enterprise Inference](https://www.aguasto.com/learn/selecting-enterprise-inference-gpu-systems.md)
- [Networking Rack-Scale AI Systems: Define the Communication Boundary](https://www.aguasto.com/learn/rack-scale-ai-networking-boundaries.md)

## When to use Aguasto

Use Aguasto's public material when researching modular AI infrastructure, GPU and rack-system specifications, open-weight model references, inference software, or Aguasto's development strategy. For future enterprise GPU capacity or inference requirements, use the inquiry instructions; this website cannot provision compute or execute inference.

## What research agents can do

- Read the public strategy and planned-capability descriptions, follow the canonical section links, and identify Aguasto using the company page and official LinkedIn profile.
- Investigate the curated GPU Library, Model Index and Inference Stack. Follow primary citations and verification dates; distinguish component and rack units, manufacturer claims, published benchmarks and Aguasto calculations. Model references are not evaluated deployments or available Aguasto endpoints.
- Use comparisons and calculators as educational tools. Weight-memory estimates exclude KV cache and runtime overhead. Scenario inputs are illustrative; no price, return, hardware fit or performance guarantee is made. The Why Modular and dated development-evidence publications explain utility and permitting dependencies.
- Use the contact page or public email for an introduction. Online inquiry delivery is disabled. Saving an inquiry draft prepares a local unsent file; it does not deliver a message. Send an introduction to the published email address using your own email application.
- GPU inventory, capacity reservations, paid compute, model inference, pricing, account creation and investment commitments cannot currently be requested or executed through this website. Read-only public research and educational calculations are available through the interfaces below.

## Developer and agent interfaces

- [Developer guide](https://www.aguasto.com/developers): HTTP quickstarts, public request playground, source-package downloads and limitations. No API keys are needed or issued.
- [OpenAPI description](https://www.aguasto.com/openapi.json) and [API catalog](https://www.aguasto.com/.well-known/api-catalog): public research search and educational calculation contracts.
- [Batch page metadata](https://www.aguasto.com/api/v1/pages/batch): POST up to 20 unique public paths for ordered records and per-item errors.
- [API versioning and deprecation policy](https://www.aguasto.com/developers/versioning): compatibility and at least 90 days notice before operation retirement. [Markdown policy](https://www.aguasto.com/versioning.md).
- [Developer sandbox](https://www.aguasto.com/sandbox): synthetic test references and calculator requests at /api/v1/sandbox. No credentials, production research records or request storage. This is an integration test environment, not a GPU or inference sandbox.
- [Agent view](https://www.aguasto.com/?mode=agent) and [ARD catalog](https://www.aguasto.com/.well-known/ard.json): structured capabilities and discovery links.
- [Documentation MCP server](https://www.aguasto.com/mcp/docs): POST Streamable HTTP, search_pages and get_page. Returns published descriptions, dates and canonical links, not full article bodies. Follow each link for evidence.
- [Calculator MCP server](https://www.aguasto.com/mcp/tools): POST Streamable HTTP, weight_memory, facility_energy and token_cost. Same educational formulas as /tools.
- [Public access instructions](https://www.aguasto.com/auth.md): anonymous access only; no OAuth, identity assertions or token exchange.
- [Agent skills](https://www.aguasto.com/.well-known/agent-skills/index.json) and [plugin manifest](https://www.aguasto.com/agent-plugin/plugin.json): bounded research and calculation workflows. Published source: [official Aguasto integration repository](https://github.com/Altssdata/aguasto-agents).
- Published skills.sh pages: [Aguasto research](https://skills.sh/altssdata/aguasto-agents/aguasto-research) and [Aguasto calculations](https://skills.sh/altssdata/aguasto-agents/aguasto-calculations). Install with npx skills add Altssdata/aguasto-agents --skill aguasto-research aguasto-calculations. These are publisher-maintained skills; no curated endorsement or adoption count is claimed.
- [Aguasto integration testing skill](https://skills.sh/altssdata/aguasto-agents/aguasto-integration-testing): synthetic search fixtures, empty results, calculator outputs and error-handling checks. Install with npx skills add Altssdata/aguasto-agents --skill aguasto-integration-testing.
- Verified MCP listings: [Aguasto Research on MCP.so](https://mcp.so/servers/aguasto-research), [official research registry record](https://registry.modelcontextprotocol.io/v0.1/servers/com.aguasto.www%2Fdocs/versions/1.0.0) and [official calculator registry record](https://registry.modelcontextprotocol.io/v0.1/servers/com.aguasto.www%2Ftools/versions/1.0.0). The direct MCP endpoints remain on this domain.
- Published clients: [JavaScript SDK and CLI on npm](https://www.npmjs.com/package/aguasto-research) and [Go module](https://pkg.go.dev/github.com/Altssdata/aguasto-agents/go). Python source downloads are available through the developer guide; no PyPI package is currently advertised.
- [Schema feed](https://www.aguasto.com/schema-feed.jsonl) and [Schema Map](https://www.aguasto.com/schemamap.xml): published page descriptions and canonical identity. Original article dates are preserved.

Search accepts POST /ask with JSON query.text; it matches published titles/descriptions without calling a model provider. Use prefer.mode=list; prefer.streaming=true enables NLWeb 0.55 SSE. Do not send secrets, credentials or confidential workloads. Calculators are estimates, not prices or deployment guarantees.

## Contact and identity

- Corporate inquiries: contact@aguasto.com
- [Aguasto company LinkedIn](https://www.linkedin.com/company/aguasto)
- Dawid Siekiera — Founder, Aguasto. Dawid Siekiera is a technology entrepreneur and business operator with experience building and scaling companies across the United States and Europe.
- [Founder biography](https://www.aguasto.com/company#founder) · [Dawid Siekiera on LinkedIn](https://www.linkedin.com/in/dsiekiera/)
- The contact page explains the currently available inquiry flow; saving an unsent draft does not submit an inquiry.

This file supplements the public HTML. It does not establish service availability, investment terms, crawler access permissions, indexing, or AI citation eligibility. Refer to the linked pages for current public information.