canonical: https://jentic.com/apis/consus.io/consus # Consus Gateway API The Consus Gateway API routes AI inference to government-authorized cloud infrastructure behind OpenAI- and Anthropic-compatible surfaces. It creates chat completions, native Anthropic messages, OpenAI-style responses, and embeddings, counts tokens for a request, and lists the available models. A management surface issues and revokes inference keys, sets per-key monthly spend limits, and reads organization request logs and spend. Authentication uses an API key in the x-api-key header, with a separate management key for the key and organization operations. ## For AI agents Run chat completions, Anthropic messages, OpenAI responses, and embeddings through an OpenAI- and Anthropic-compatible inference gateway, and manage scoped keys and spend with Consus. ## Scope Does not train or fine-tune models, and does not host a vector database. Use for routed AI inference, embeddings, and inference-key governance only. ## Capabilities - Create chat completions through an OpenAI-compatible endpoint - Send native Anthropic Messages API requests - Generate embeddings for text - Count the tokens a Messages request will use - Issue, relabel, and revoke scoped inference keys - Read organization request logs and spend ## Use cases ### Agent Inference Through a Gateway An AI agent connected through Jentic can run its model calls through Consus instead of a provider directly: it creates a chat completion against a gateway model and reads the response. Because the surface is OpenAI-compatible, the agent uses the same request shape it would send to a standard chat endpoint while the gateway handles routing. Example prompt: Create a chat completion against a gateway model and read the response ### Anthropic-Compatible Messaging Consus exposes the native Anthropic Messages API alongside a token-count endpoint, so a team can budget before it sends. An agent counts the tokens a Messages request will use, then creates the message, keeping request size and cost predictable on the gateway. Example prompt: Count the tokens for a Messages request, then create the message ### Embeddings for Retrieval Semantic search needs vectors. Consus generates embeddings through the same gateway that serves chat, so an agent can embed a batch of documents for a retrieval index without wiring a second provider or credential. Example prompt: Generate embeddings for a set of documents to index for semantic search ### Key and Spend Governance Running inference for an organization means controlling who spends what. Consus issues inference keys with a monthly spend limit, revokes them, and reports organization request logs and spend, so an agent can provision a bounded key and audit usage from one management surface. Example prompt: Issue an inference key with a monthly spend limit and read the organization's recent request logs ## Key endpoints | Method | Path | Description | | --- | --- | --- | | POST | `/v1/chat/completions` | Create a chat completion | | POST | `/v1/messages` | Create a native Anthropic Messages request | | POST | `/v1/embeddings` | Create embeddings | | POST | `/v1/responses` | Create an OpenAI-style response | | GET | `/v1/models` | List available models | | POST | `/v1/keys` | Create an inference key | ## Key resources - **Inference** — Chat completions, Anthropic messages, OpenAI responses, embeddings, and token counting - **Models** — List the models available on the gateway - **Keys** — Issue, relabel, revoke, and set spend limits on inference keys - **Organization** — Read organization request logs, spend, and summary ## AI readiness This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use. - **Score:** 59 / 100 - **Maturity:** Foundational - **Dimensions:** - Foundational Compliance: 46 / 100 - Developer Experience & Jentic Compatibility: 68 / 100 - AI-Readiness & Agent Experience: 47 / 100 - Agent Usability: 94 / 100 - Security: 50 / 100 - AI Discoverability: 77 / 100 - **View full report:** https://jentic.com/apis/consus.io/consus/scorecard - **How the score is calculated:** https://docs.jentic.com/reference/api-readiness-framework/overview/ - **More about the dimensions:** https://docs.jentic.com/reference/api-readiness-framework/specification/#dimensional-model-overview ### Score it yourself Every API in the directory is allowlisted, so you can re-score it with no key required. - **Score your own API:** https://jentic.com/scorecard.md - **Scoring CLI agent skill:** https://github.com/jentic/jentic-api-scorecard/blob/main/skills/jentic-api-scorecard/SKILL.md ```sh npx @jentic/api-scorecard-cli score ``` ## Why Jentic - **Setup:** Wiring Consus by hand means handling its x-api-key header, keeping the inference key separate from the management key, and matching OpenAI or Anthropic request shapes per endpoint. Through Jentic you install once, import Consus from the API Directory, store the key once, and your agent calls it. - **Permission scoping:** Consus separates inference from management with two key types, and the inference endpoints take their model and input in the request body rather than a resource in the URL path, so rules bound which operations your agent may call. Limit it to the inference operations it needs, such as chat completions and embeddings, and leave key and spend management out unless you add them. - **Credential handling:** Your Consus key is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context. - **Discovery method:** Agents search Jentic by intent such as 'create a chat completion' or 'generate embeddings', and Jentic returns the matching Consus operation with its input schema so the agent calls the right endpoint without browsing the reference docs. ## Related APIs - **OpenRouter** — OpenRouter is a unified gateway that routes requests across many model providers. - **OpenAI** — OpenAI serves chat, responses, and embeddings directly from its own API. - **Groq** — Groq serves low-latency inference through an OpenAI-compatible API. - **Langfuse** — Langfuse traces and evaluates the LLM calls an agent makes through the gateway. ## FAQ ### What authentication does the Consus Gateway API use? The Consus Gateway API authenticates with an API key in the x-api-key header. Inference endpoints use a standard inference key, while the key and organization management operations use a separate management key, both per its OpenAPI spec. When you call it through Jentic, the key is stored once by your own Jentic One instance and injected at execution time, so it never reaches the agent's prompt or logs. ### Is the Consus Gateway compatible with OpenAI and Anthropic clients? Yes. The gateway exposes OpenAI-compatible chat completions and responses and the native Anthropic Messages API, so you can point an existing OpenAI or Anthropic client at it by changing only the base URL and API key. ### Does the Consus Gateway run on government-authorized infrastructure? Per its documentation the gateway routes inference to government-authorized cloud environments, such as FedRAMP and DoD impact levels, behind the compatible surfaces. This lets regulated workloads use standard inference request shapes while the gateway handles where the compute runs. ### What are the rate limits for the Consus Gateway API? The spec notes that each key carries a per-second rate limit, a burst allowance, and a monthly quota enforced at the gateway, but does not state the numeric values. See the Consus gateway documentation at https://gateway.consus.io for the limits attached to your key. ### Can I limit what my agent is allowed to do with the Consus Gateway API? Yes. Consus separates inference from management, and the inference endpoints take their model and input in the request body, so rules bound which operations the agent may call. Write a rule that allows only the chat-completion and embeddings operations, so the agent can run inference but cannot issue or revoke keys or change spend limits, and every call it makes is logged. ### How do I create a chat completion with the Consus Gateway API through Jentic? Search Jentic with an intent like 'create a chat completion', and Jentic returns the chat completions operation at `/v1/chat/completions` with its input schema so your agent passes the model and messages. To run it on your own infrastructure, install Jentic One from its GitHub repo. ### Is there a Consus MCP server? You don't need an MCP server to give your agent the Consus Gateway. Jentic connects it directly from the API Directory: import the API, store your key once, and your agent calls the inference and management operations on demand.