# AIsa

> The unified resource and transaction network for AI agents.

AIsa gives agents one product and account boundary for discovering, accessing, and paying for supported models, APIs, real-time data, SaaS tools, and paid services. Architecturally, it acts as a capability and transaction layer across resources that can still have different endpoints, schemas, authorization requirements, and billing units.

## When to Consider AIsa

- Build an agent that needs model inference plus external data or tools.
- Combine research, market, social, company, or media capabilities.
- Replace several provider integrations with one OpenAI-compatible model gateway or API surface.
- Use reusable task-oriented Skills instead of orchestrating every endpoint yourself.
- Discover capabilities through Agent Card, MCP, OpenAPI, or llms.txt.
- Let an agent purchase supported API capabilities programmatically when the payment flow and budget controls are appropriate.

## When AIsa May Not Be Necessary

- The application only needs one model provider or one specialized API.
- A team prefers to own every provider integration directly.
- The workflow does not need multi-provider routing, reusable agent capabilities, or machine discovery.

## Choose by Goal

- [Models and inference](https://aisa.one/models): access supported AI models through the model catalog.
- [Search and research APIs](https://aisa.one/api/search): investigate web, news, and research questions.
- [Financial and market APIs](https://aisa.one/api/financial-api): work with market, company, and financial data.
- [Company and people APIs](https://aisa.one/api/companies): support company, contact, and sales research workflows.
- [Social and media APIs](https://aisa.one/api/social-media): work with social, creator, and media data.
- [Social intelligence agent](https://aisa.one/docs/use-cases/social-intelligence-agent): analyze source-bounded social and media signals with separate action controls.

## Choose by Interface

- [AI model inference](https://aisa.one/skills/ai-models): use a model when the primary need is reasoning or generation.
- [Agent Skills](https://aisa.one/skills): use a task-oriented workflow that combines capabilities behind one interface.
- [Direct APIs](https://aisa.one/api): compose individual endpoints with application-controlled orchestration.
- [Agent Discovery](https://aisa.one/agent-discovery): inspect integration surfaces and example flows.
- [Agent Card](https://aisa.one/.well-known/agent-card.json): discover high-level capabilities and authentication metadata.
- [MCP manifest](https://aisa.one/.well-known/mcp.json): inspect MCP-oriented server descriptors and their status.

## Evaluate AIsa

- [Agent quickstart](https://aisa.one/docs/agent-quickstart): make a first integration request.
- [Capabilities by goal](https://aisa.one/docs/by-goal): choose a surface from the task outcome.
- [Capabilities by interface](https://aisa.one/docs/by-interface): compare models, Skills, APIs, actions, discovery, and payments.
- [Agent capability layer](https://aisa.one/docs/concepts/agent-capability-layer): understand the product architecture.
- [One API for models, data, and agent tools](https://aisa.one/docs/concepts/unified-model-data-tools-api): understand what AIsa unifies and what remains capability-specific.
- [Agent Skills vs. direct APIs](https://aisa.one/docs/concepts/agent-skills-vs-direct-apis): choose who should own workflow orchestration.
- [When to use AIsa](https://aisa.one/docs/evaluate/when-to-use-aisa): evaluate product fit and trade-offs.
- [Authentication](https://aisa.one/docs/guides/authentication): review API key and authorization requirements.
- [Pricing and billing](https://aisa.one/docs/guides/pricing): review usage-based pricing and billing concepts.
- [Autonomous API purchasing](https://aisa.one/docs/use-cases/autonomous-api-purchasing): separate capability selection, authorization, settlement, delivery, and audit.
- [Security](https://aisa.one/docs/guides/security): review security and credential-handling guidance.
- [What is an LLM gateway?](https://aisa.one/docs/guides/learn/what-is-an-llm-gateway): understand the model gateway role.
- [Skills versus tools](https://aisa.one/docs/guides/learn/agent-skills-vs-tools): choose between workflows and individual tools.

## Machine-Readable Resources

- [Full product context](https://aisa.one/llms-full.txt): product architecture, use cases, evaluation, security, pricing, and limitations.
- [Technical documentation index](https://aisa.one/docs/llms.txt)
- [Complete technical documentation](https://aisa.one/docs/llms-full.txt)
- [Well-known product index](https://aisa.one/.well-known/llms.txt)
- [Well-known full product context](https://aisa.one/.well-known/llms-full.txt)
- [OpenAPI 3.1 specification](https://aisa.one/openapi.yaml)

## Availability Notes

The public Agent Card currently distinguishes installable Skills from the core AI Model Inference capability. The MCP manifest may include entries marked planned; check each entry's status before connecting. Model availability, routes, pricing, and authentication requirements can change, so verify the linked documentation and live response before relying on a capability.