Page role
Resource Library is the official source list.
Use this page when you need official pricing pages, cloud references, model documentation and trusted support links. For interpretation and decisions, use AI Infrastructure Hub. For cost examples and warnings, use Prices and Cost Planning.
AI Infrastructure Hub Prices Cost PlanningAI Builder Resource Library 2026
This GPUJet resource library collects the most useful official links, pricing pages, infrastructure references, AI agent resources and troubleshooting paths for people building AI projects in 2026. Use it as a practical starting point before choosing a model API, VPS, GPU cloud provider or AI automation stack.
The goal is simple: instead of searching scattered sources every time, builders can start here, open the official documentation, compare costs, then follow the right GPUJet tutorial for the next step.
Model API pricing and documentation
| Resource | Use it for | Official link |
|---|---|---|
| OpenAI API pricing | Check input tokens, cached input, output tokens, batch pricing, realtime, audio and tool/container pricing. | OpenAI pricing |
| Claude API pricing | Check Anthropic model pricing, cache writes, cache hits, batch pricing and model tiers. | Claude pricing |
| Gemini API pricing | Check free tier, paid tier, context caching, batch API, image, audio and Google Search grounding pricing. | Gemini API pricing |
| DeepSeek API pricing | Check V4 Flash, V4 Pro, cache-hit pricing, discounts and balance deduction rules. | DeepSeek pricing |
| AWS Bedrock pricing | Check enterprise model access, managed foundation model usage and AWS billing details. | AWS Bedrock pricing |
| AWS Bedrock AgentCore pricing | Check agent runtime, memory, gateway, browser/code tools and consumption-based agent infrastructure. | AgentCore pricing |
Cloud, VPS and GPU pricing
| Resource | Use it for | Official link |
|---|---|---|
| Hostinger OpenClaw | Beginner-friendly OpenClaw positioning, managed setup and AI agent entry point. | Hostinger OpenClaw |
| Hostinger OpenClaw VPS Docker template | OpenClaw VPS deployment path, Docker template, recommended server path and setup context. | OpenClaw VPS template |
| Hostinger VPS pricing | Check current VPS plans before running small bots, dashboards, OpenClaw tests or API-based apps. | Hostinger VPS pricing |
| DigitalOcean Droplet pricing | Check normal VPS pricing for APIs, dashboards, small apps and background workers. | DigitalOcean Droplets |
| DigitalOcean GPU Droplets | Check GPU hourly rates, GPU memory, regions and high-end configurations. | DigitalOcean GPU pricing |
| RunPod GPU pricing | Check GPU pods, serverless GPU and AI experiment pricing. | RunPod pricing |
| Vast.ai pricing | Check marketplace GPU rates and compare host quality, storage, reliability and availability. | Vast.ai pricing |
GPUJet internal learning path
- Start Here — the beginner entry point for choosing a first AI project path.
- AI Agent — understand agents as workflows with inputs, tools, guardrails, logs and outputs.
- Cloud — choose between normal hosting, VPS, API-first AI, GPU cloud and enterprise cloud.
- Prices — compare model API prices, GPU hourly math, hosting costs and cost warnings.
- OpenClaw for Beginners — learn OpenClaw as a practical AI agent workflow path.
- AI API Cost Control Tutorial — control token usage, limits, retries and billing risk.
- GPU Cloud Decision Guide — decide when GPU cloud is worth renting.
- AI Agent Safety Checklist — apply approval, logs, backups, limits and rollback thinking.
Agent tools and workflow builders
Beginners should not choose an agent tool only because it is popular. First define the workflow: input, trigger, tools, permissions, approval, logs and output. Then choose the simplest tool that can safely run that workflow.
| Tool type | Examples | Best beginner use |
|---|---|---|
| AI agent workflow | OpenClaw-style workflows | Draft-only workflows, safe testing and beginner agent concepts. |
| Workflow automation | n8n-style automation | Triggers, webhooks, app actions and repeatable business processes. |
| Visual LLM builders | Flowise, Langflow-style tools | Prototypes, RAG flows, prompt chains and visual experimentation. |
| Developer agent frameworks | LangGraph-style architecture | Advanced stateful agents, custom apps, multi-step workflows and production control. |
| Local model tools | Ollama-style local inference | Learning, privacy experiments and small-model testing before renting GPU cloud. |
Troubleshooting and community research
Official documentation tells you how a tool should work. Community discussions show where users actually get stuck. Use both, but never share API keys, SSH keys, admin passwords, private URLs, payment details or screenshots containing sensitive information.
- Use official docs first for pricing, setup steps and supported features.
- Use GitHub issues for known bugs, release changes and installation errors.
- Use Stack Overflow-style searches for exact error messages, Docker issues, ports, Linux services and deployment problems.
- Use Reddit or forum discussions only as clues, not as final technical truth.
- Use YouTube walkthroughs for visual orientation, then verify commands against current official documentation.
Before you spend money checklist
- Define the exact AI task: draft, classify, summarize, retrieve, generate, monitor or automate.
- Decide whether the project needs to call a hosted model or run a model directly.
- Estimate one test run, one daily scenario and one monthly scenario.
- Set API limits, budgets or alerts before sharing the workflow.
- Keep the first version draft-only or approval-required.
- Create a rollback path: disable key, stop server, restore backup or disconnect webhook.
- Open the official pricing page before payment.
GPUJet rule: the best setup is not the most powerful one. The best setup is the smallest one you understand, can afford, can monitor and can safely roll back.