Install Jentic One Beta

Connect the LLMIntel API to your agent

Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the LLMIntel API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.

Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.

Step 1: Jentic One Host machine

# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fllmintel.ai%2Fllmintel" | sh

Step 2: Agent machine

# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fllmintel.ai%2Fllmintel" | sh
jentic register       # connects your agent to your Jentic One instance

Jentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.

Capabilities

What an agent can do with LLMIntel API.

List tracked AI models filtered by provider (openai, anthropic, azure, bedrock, google, mistral, cohere) and lifecycle state via listModels

Look up one model's canonical lifecycle state by id, alias, or bare model string with getModel

Poll lifecycle change events newest first with the since timestamp filter to catch deprecations and retirements via listEvents

Compare two to six models side by side on lifecycle and cost with compareModels and the comma-separated models parameter

GET STARTED

Start building with LLMIntel API

Explore with Jentic One

View OpenAPI Document

Use Cases

Patterns agents use LLMIntel API for, with concrete tasks.

★ Agent-driven model deprecation checks through Jentic

A coding agent working on an LLM application asks Jentic for the LLMIntel operation that reports model lifecycle state, then calls getModel for each model string found in the codebase. If a model comes back as deprecated, retiring, or retired, the agent flags the call site and proposes a replacement. The bearer key used for comparison and telemetry stays in your self-hosted Jentic One instance.

Search Jentic for 'check whether an AI model is deprecated', call getModel with id 'gpt-4o', and report the returned lifecycle state

Incremental lifecycle event polling

A scheduled job calls listEvents with the since parameter set to the timestamp of its previous run, optionally filtered by provider and target state, and receives only new lifecycle changes, newest first. Each lifecycle record links to its primary provider source, so the job can attach the announcement to an internal alert or ticket.

Call listEvents with provider 'anthropic', state 'retiring', and since '2026-11-01T00:00:00Z', then summarize each returned event

Pre-migration model comparison

Before swapping one model for another, a developer calls compareModels with a comma-separated list such as gpt-4o and anthropic/claude-opus-4-1. LLMIntel resolves canonical ids, aliases, or bare model names into one comparison matrix across lifecycle and cost; performance fields such as speed and latency are filled only for paid keys.

Call compareModels with models 'gpt-4o,anthropic/claude-opus-4-1' and summarize the lifecycle and cost differences

Runtime LLM spend tracking

An application posts batches of metadata-only usage records to ingestTelemetry. LLMIntel hour-truncates each timestamp, resolves and prices the model at ingest so historic spend never shifts, and upserts the records into per account, environment, model, and hour rollups, giving a live cost view without sending prompt or response content.

Call ingestTelemetry with a records array containing the last hour of model usage for the production environment and confirm the records were accepted

Key Endpoints

6 endpoints — the llmintel api tracks lifecycle events for ai models across providers such as openai, anthropic, azure, aws bedrock, and google, and turns runtime usage telemetry into a live cost and usage view of the models your application calls.

METHOD

PATH

DESCRIPTION

GET

/v1/models

List models with lifecycle state

GET

/v1/models/{id}

Get one model

GET

/v1/events

List lifecycle events

GET

/v1/compare

Compare models side by side

POST

/v1/telemetry

Ingest runtime usage

GET

/v1/status

Freshness and SLA status

GET

/v1/models

List models with lifecycle state

GET

/v1/models/{id}

Get one model

GET

/v1/events

List lifecycle events

GET

/v1/compare

Compare models side by side

POST

/v1/telemetry

Ingest runtime usage

GET

/v1/status

Freshness and SLA status

Why Jentic?

What agents get from Jentic-routed access to this vendor.

Setup

Wiring LLMIntel by hand means telling which lifecycle reads work anonymously and which calls need the bearer key, and handling its X-RateLimit headers yourself. Through Jentic you install once, import LLMIntel from the API Directory, store the key once, and your agent calls it.

Permission scoping

LLMIntel puts the model id in the URL path for single-model lookups, but most calls carry their filters as query parameters or a request body, so rules bound which operations your agent may call rather than which models. You choose the operations it may call, so telemetry ingest is not included unless you add it.

Credential isolation

Your LLMIntel 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.

Intent-based discovery

Agents search Jentic by intent such as 'is this model deprecated' or 'compare two LLMs', and Jentic returns the matching LLMIntel operation with its input schema so the agent calls the right endpoint without browsing the reference docs.

FAQs

Specific to using LLMIntel API through Jentic.

What authentication does the LLMIntel API use?

Per its OpenAPI spec, the LLMIntel API uses an HTTP bearer token sent as `Authorization: Bearer` with a key created from your LLMIntel dashboard. The lifecycle reads (listModels, getModel, listEvents) accept anonymous calls or a key, and getStatus needs no key; compareModels and ingestTelemetry require the key, and telemetry ingest expects an account-scoped key. Through Jentic the key is stored encrypted by your own Jentic One instance and injected at call time.

Can I check whether a model is deprecated with the LLMIntel API?

Yes. Call GET /v1/models/{id} with a canonical id, an alias, or the bare model string your code uses, and read the returned lifecycle state, which is one of announced, active, legacy, deprecated, retiring, or retired. To watch for changes over time, poll GET /v1/events with the since parameter.

What are the rate limits for the LLMIntel API?

The OpenAPI spec does not state numeric limits. Responses carry X-RateLimit-Limit, X-RateLimit-Remaining, and X-RateLimit-Reset headers, and exceeding the limit returns a 429 with a Retry-After header. Anonymous callers share a stricter per-IP budget, and sending a key raises it to that key's tier.

How do I poll for new model retirements with the LLMIntel API through Jentic?

Install Jentic One with `curl -fsSL https://raw.githubusercontent.com/jentic/jentic-one/main/tools/install.sh | sh`, add the LLMIntel API from the Jentic API Directory, and have your agent search for 'list AI model lifecycle events'. Jentic returns listEvents (GET /v1/events) with its schema, and the agent passes state 'retired' and a since timestamp to fetch only new changes. To run it on your own infrastructure, install Jentic One from its GitHub repo.

Is there an LLMIntel MCP server?

You don't need an MCP server to give your agent the LLMIntel API. Jentic connects it directly from the API Directory: import it, store your credential once, and your agent calls it. Operations are discovered on demand, so no extra tool definitions sit in your agent's context.

Can I limit what my agent is allowed to do with the LLMIntel API?

Yes. For a read-only lifecycle checker, write a rule that allows listModels, getModel, listEvents, and getStatus. The agent can then look up model states and poll events, while ingestTelemetry is outside the allowed set, so it cannot write usage records to your account, and every call it makes is logged.

Does the LLMIntel API receive my prompts when I send telemetry?

No. POST /v1/telemetry takes metadata-only usage records with no prompt or response content. The server hour-truncates each timestamp, prices the model at ingest, and upserts the record into per account, environment, model, and hour rollups.

LLMIntel API

AI ReadinessAI-Aware (B)65/100

See full scorecard

Official vendor OpenAPI document · agent-readyDeveloper ToolsMonitoring Observabilityhttp bearer (optional on lifecycle reads)6 EndpointsREST

For Agents

Check whether an AI model is active, deprecated, retiring, or retired, poll lifecycle change events by provider, compare models on lifecycle and cost, and send usage telemetry to LLMIntel for cost tracking.

Use for: Check whether the model my app calls is being deprecated, List all Anthropic models that are in the retiring state, Find all lifecycle events since last Monday for OpenAI models, I want to compare the cost of two models before switching

Not supported: Does not run model inference, proxy LLM requests, or store prompts and responses. Use for AI model lifecycle tracking, model comparison, and usage cost telemetry only.

The LLMIntel API tracks lifecycle events for AI models across providers such as OpenAI, Anthropic, Azure, AWS Bedrock, and Google, and turns runtime usage telemetry into a live cost and usage view of the models your application calls. Lifecycle reads cover the model list, single-model detail, a newest-first event feed with a since filter for incremental polling, and a freshness status check, and they work without an API key. Authenticated calls compare two to six models side by side across lifecycle, cost, and performance, and ingest metadata-only usage records that are priced at ingest and rolled up per account, environment, model, and hour.

Jentic One on Github View OpenAPI Document

Show dimension breakdown

Foundational Compliance

Base layer of spec validity and structural soundness.

Grade: A+Signals: 4

98%

Lint Results

Aggregated quality score from linter diagnostics, weighted by severity.

100%

Resolution Completeness

Percentage of `$ref` references that resolve successfully.

100%

Specification Validity

Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).

75%

Structural Integrity

Structural correctness score based on schema issues using logarithmic dampening.

Developer Experience & Jentic Compatibility

Clarity, completeness, and ingestion readiness for developers and tooling.

Grade: BSignals: 4

Example Density

How richly the API is illustrated with examples.

100%

Example Validity

Percentage of examples that conform to their schemas.

50%

Response Coverage

Percentage of operations with complete response definitions (success, client error, server error).

100%

Tooling Readiness

Health of API ingestion, bundling, and resolution within Jentic pipelines.

AI-Readiness & Agent Experience

Semantic breadth, depth, and agent comprehension for AI systems.

Grade: DSignals: 4

35%

Description Coverage

Coverage of descriptions across API elements.

Error Standardization

Coverage of RFC 9457 Problem Details for error responses.

100%

OperationId Quality

Coverage, uniqueness, and casing consistency of operationIds for AI inference.

46%

Summary Coverage

Coverage of summaries across operations/tags/info.

Agent Usability

Functional utility, complexity comfort, and AI orchestration readiness.

Grade: A+Signals: 1

94%

Complexity Comfort

Agent comfort level based on API operational and structural complexity.

Security

Trust, risk posture, and security compliance.

Grade: B-Signals: 1

60%

Authentication Strength

Average quality of security schemes based on authentication method strength (weakest link for OAuth2).

AI Discoverability

Findability, semantic richness, and reasoning readiness.

Grade: A-Signals: 1

71%

Descriptive Richness

Clarity and depth of descriptions across API elements.

Connect the LLMIntel API to your agent

Capabilities

Start building with LLMIntel API

Use Cases

Key Endpoints

Why Jentic?

Related APIs

FAQs

LLMIntel API

Jentic AI Readiness Score

LLMIntel API

Foundational Compliance

Lint Results

Resolution Completeness

Specification Validity

Structural Integrity

Developer Experience & Jentic Compatibility

Example Density

Example Validity

Response Coverage

Tooling Readiness

AI-Readiness & Agent Experience

Description Coverage

Error Standardization

OperationId Quality

Summary Coverage

Agent Usability

Complexity Comfort

Security

Authentication Strength

AI Discoverability

Descriptive Richness