ubq.fi valid

UbiquityOS AI Gateway (AI.UBQ.FI)

OpenAI-compatible AI inference service and developer API for the UbiquityOS ecosystem.

Their llms.txt

The beginning of the file, from our validated copy saved 2026-10-26T05:13:22.516Z. Cached copy · Source on ubq.fi

# UbiquityOS AI Gateway (AI.UBQ.FI)

OpenAI-compatible AI inference service and developer API for the UbiquityOS ecosystem.

Canonical service URL: https://ai.ubq.fi
OpenAI-compatible base URL: https://ai.ubq.fi/v1
Authentication: Authorization: Bearer <UOS_AI_TOKEN>

## Authoritative resources

- OpenAPI 3.1 contract: https://ai.ubq.fi/openapi.json
- Developer portal and quickstart: https://ai.ubq.fi/developers
- Complete agent instructions: https://ai.ubq.fi/llms-full.txt
- Supported models: GET https://ai.ubq.fi/v1/models
- Responses: POST https://ai.ubq.fi/v1/responses
- Chat completions: POST https://ai.ubq.fi/v1/chat/completions
- Decisions: POST https://ai.ubq.fi/v1/decisions

## When to use this service

Use the UbiquityOS AI Gateway when an approved UbiquityOS application or autonomous agent needs
OpenAI-compatible inference, runtime model discovery, streaming Responses or Chat Completions, or documented function
tools. Start with the OpenAPI contract, query `/v1/models` before choosing a model, and call the `/v1` endpoints with an
approved bearer token.

Do not use this service to create credentials, administer API keys, access private admin or provider-health routes, or
guess model names. The browser chat playground is an authenticated test surface for a token already issued to the user;
it is not a self-service key issuer. If no approved token is available, stop and ask the authorized operator for access.

## Autonomous use

1. Read this file and the linked OpenAPI contract.
2. Obtain `UOS_AI_TOKEN` from the environment or an approved secret manager.
3. Configure an OpenAI-compatible client with base URL `https://ai.ubq.fi/v1` and the token as its API key.
4. Query `/v1/models` and select a returned model. Do not assume that an OpenAI public model alias is available.
5. Use `/v1/responses` or `/v1/chat/completions` with the official OpenAI request shape. Use `/v1/decisions` only for a
   model whose advertised `supported_endpoints` includes it.
6. On a `429`, parse the JSON error and honor `Retry-After` when present; use `RateLimit` fields when they are present.

Never print, log, persist, transmit elsewhere, or commit the token. If no token is available, tell the user that service
authentication is required. Do not request that the user paste a token into chat.

Related sites

Explore similar publishers →

Category suggestions: other listed sites in AI and machine learning. These share a category. See the whole category →

Machine-readable: GET /v1/resolve?q=ubq.fi · GET /v1/domain/ubq.fi · this page as Accept: application/json or text/markdown · MCP lookup_domain

Automatically discovered from public files. Inclusion does not imply endorsement or publisher participation.

Claim my listing on Agent Web Map for ubq.fi:
1. POST https://agentwebmap.com/v1/sites/ubq.fi/claims and keep claim_id and claim_secret private.
2. Publish the verification line "AgentWebMap verification: <token>" (token from step 1) in https://ubq.fi/llms.txt, or use the DNS, meta-tag or well-known method the response describes. This only verifies the listing; your website and domain remain yours.
3. POST https://agentwebmap.com/v1/sites/ubq.fi/verify with {"claim_id": "...", "claim_secret": "..."}.
4. Save the api_key from the answer; it manages the listing (PATCH https://agentwebmap.com/v1/sites/ubq.fi).
MCP equivalent: tools begin_ownership_claim, then verify_domain, at https://agentwebmap.com/mcp. Docs: https://agentwebmap.com/docs#verify

Corrections or removal requests: contact the directory.