# AI Model Watch (npm · @aimodelwatch/mcp)

Prices, context windows, lifecycle and deprecation dates for 270+ AI/LLM models. Daily, no key.

- Trust score: 66/100 (medium)
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-09-20

## Components

- npm · `@aimodelwatch/mcp`: 66/100 (this document), [markdown](https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp.md), [page](https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp)

## Channel facts

- Registry: `npm`
- Package: `@aimodelwatch/mcp`
- Version: `2.0.1`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-09-20.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - No install/post-install scripts declared.
  - 0 of 3 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 19/100
  - Repository check failed: the declared repository URL returned HTTP 404.
  - Provenance check failed: no build-provenance attestation is published.
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 0 days ago).
  - Security-disclosure policy not yet verified: we couldn't inspect the source repository.
- **Schema Quality & AI Usability**: 86/100
  - 100% of prompts and resources have a non-trivial description (not blank, and not just the item's name).
  - AI-judged instruction clarity (excellent).
  - Context-footprint check failed: tool/resource definitions use about 1213 tokens (~110/item across 11 items; 9 tools + 2 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 89/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 63% of tool parameters carry a description.
  - Structured output schemas are declared (100% of tools); any adoption earns full credit.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 9 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 11 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

**Unverified: 1 category.** A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.

## Install

### How do I install the AI Model Watch MCP server?

AI Model Watch runs locally as an npm package, launched with npx -y @aimodelwatch/mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add khavel-aimodelwatch -- npx -y @aimodelwatch/mcp
```

### Cursor

```json
{
  "mcpServers": {
    "khavel-aimodelwatch": {
      "command": "npx",
      "args": [
        "-y",
        "@aimodelwatch/mcp"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "khavel-aimodelwatch": {
      "command": "npx",
      "args": [
        "-y",
        "@aimodelwatch/mcp"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add khavel-aimodelwatch -- npx -y @aimodelwatch/mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "khavel-aimodelwatch": {
      "type": "local",
      "command": [
        "npx",
        "-y",
        "@aimodelwatch/mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add khavel-aimodelwatch --command npx --arg -y --arg @aimodelwatch/mcp
```

### Hermes

```yaml
mcp_servers:
  khavel-aimodelwatch:
    command: "npx"
    args: ["-y", "@aimodelwatch/mcp"]
```

### Netclaw

```json
{
  "McpServers": {
    "khavel-aimodelwatch": {
      "Transport": "stdio",
      "Command": "npx",
      "Arguments": [
        "-y",
        "@aimodelwatch/mcp"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add khavel-aimodelwatch -t stdio -c npx -a -y @aimodelwatch/mcp
```

### Other

```json
{
  "mcpServers": {
    "khavel-aimodelwatch": {
      "command": "npx",
      "args": [
        "-y",
        "@aimodelwatch/mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-09-20 (score 66)

First indexed and scored.

## MCP tools (9)

### `list_models` (~202 tokens)

List models

List models from the catalog with optional filters and pagination. Returns summary rows (no notes); call get_model for the full row. Filters combine with AND.

Input parameters:

- `limit` (integer)
- `modality` (string): Substring match on the modality tags, e.g. 'text', 'image', 'audio', 'video', 'pdf', 'code'
- `offset` (integer)
- `open_weight` (boolean): true = open-weight models only, false = closed only
- `provider` (string): Provider name, case-insensitive substring: 'openai', 'anthropic', 'google', 'mistral'...
- `query` (string): Case-insensitive text search over id, name and api_string
- `status` (string): Lifecycle status as the provider states it: 'ga', 'preview', 'beta', 'deprecated', 'retired'. 'active' is shorthand for ga+preview+beta (still callable).

Output parameters:

- `count` (integer)
- `models` (array)
- `next_offset`
- `offset` (integer)
- `total` (integer)
- `updated` (string): Date the catalog last changed (feed `updated`)

### `get_model` (~90 tokens)

Get model

Full catalog row for one model, looked up by id (slug) or api_string, case-insensitive. Includes notes, knowledge cutoff and the official source_url. When not found, returns found=false plus up to 5 similar ids.

Input parameters:

- `id` (string, required): Model id or api_string, e.g. 'claude-opus-4-8' or 'gpt-5.6'

Output parameters:

- `found` (boolean)
- `model`
- `suggestions` (array)
- `updated` (string)

### `search_models` (~64 tokens)

Search models

Free-text search across id, name, provider, api_string AND the notes field (the notes carry batch prices, cache tiers, hosting availability...). For structured filtering prefer list_models.

Input parameters:

- `limit` (integer)
- `query` (string, required): All whitespace-separated terms must match

Output parameters:

- `models` (array)
- `total` (integer)
- `updated` (string)

### `compare_models` (~41 tokens)

Compare models

Side-by-side rows for 2-6 models (by id or api_string) plus blended price per 1M tokens.

Input parameters:

- `ids` (array, required)

Output parameters:

- `models` (array)
- `not_found` (array)
- `updated` (string)

### `cheapest_models` (~108 tokens)

Cheapest models

The N cheapest models by blended price (mean of input and output USD per 1M tokens). Only models with both prices published. Defaults to active models; pass status to widen.

Input parameters:

- `count` (integer)
- `modality` (string): e.g. 'text', 'image', 'embedding'
- `provider` (string): Provider name, case-insensitive substring: 'openai', 'anthropic', 'google', 'mistral'...
- `status` (string)

Output parameters:

- `models` (array)
- `updated` (string)

### `check_model_status` (~98 tokens)

Check model status

Is this model id safe to pin? Reports the provider-declared lifecycle status, the deprecation/retirement dates, days until retirement, and the stated replacement (with its row). A model absent from the catalog is reported as found=false, not as retired.

Input parameters:

- `id` (string, required): Model id or api_string, e.g. 'gpt-4o', 'claude-3-5-sonnet-20241022'

Output parameters:

- `api_string` (string|null)
- `days_to_retirement` (number|null)
- `deprecated_on`
- `found` (boolean)
- `id` (string|null)
- `lifecycle`
- `name` (string|null)
- `replacement` (string|null)
- `replacement_model`
- `retires_on`
- `source_url` (string|null)
- `status`
- `suggestions` (array)
- `updated` (string)

### `list_deprecations` (~191 tokens)

List deprecations

Every deprecation/retirement a provider has declared, with dates, days to retirement and the stated replacement. 'upcoming' = not yet retired (dated in the future, or undated); 'past' = already retired. Sorted soonest retirement first; undated rows last.

Input parameters:

- `after` (string): Keep rows retiring (or, if undated, deprecated) on/after this date
- `as_of` (string): Reference date for upcoming/past and day counts; defaults to today (UTC)
- `before` (string): Keep rows retiring (or, if undated, deprecated) on/before this date
- `limit` (integer)
- `offset` (integer)
- `provider` (string): Provider name, case-insensitive substring: 'openai', 'anthropic', 'google', 'mistral'...
- `when` (string)

Output parameters:

- `as_of` (string)
- `count` (integer)
- `deprecations` (array)
- `next_offset`
- `offset` (integer)
- `total` (integer)
- `updated` (string): Date the catalog last changed (feed `updated`)

### `models_changed_since` (~195 tokens)

Models changed since

Lifecycle events dated on/after a date, newest first: launches (released), deprecation announcements (deprecated_on) and retirements that have happened (retires_on). Built from the dated fields on each catalog row; the feed has no per-field modification timestamps, so price changes are not visible here (see aimodelwatch.dev/changelog). Retirements scheduled after today are omitted unless include_scheduled=true (use list_deprecations when=upcoming for those). Also returns the feed `updated` date so you can tell whether the catalog moved at all.

Input parameters:

- `include_scheduled` (boolean): Also list retirements dated in the future (event "retires")
- `limit` (integer)
- `provider` (string): Provider name, case-insensitive substring: 'openai', 'anthropic', 'google', 'mistral'...
- `since` (string, required): YYYY-MM-DD, inclusive

Output parameters:

- `as_of` (string)
- `count` (integer)
- `events` (array)
- `note` (string)
- `since` (string)
- `total` (integer)
- `updated` (string)

### `feed_freshness` (~56 tokens)

Feed freshness

How current is the data: the catalog `updated` date, its age in days, model/deprecation counts, providers covered, and this server's cache state. Free to call; hits the network at most once per TTL.

Output parameters:

- `as_of` (string)
- `cache` (object)
- `days_since_update` (integer)
- `deprecation_count` (integer)
- `feed_updated` (string)
- `feed_urls` (object)
- `license` (string)
- `model_count` (integer)
- `providers` (array)
- `updated` (string)

## Diagnostics

Captured diagnostic sections: Provenance, Dependencies. The full working is on the page: https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp#diagnostics

## Score history

- 2026-09-20: 66

## Common questions

### What is the AI Model Watch MCP server?

AI Model Watch is an MCP server listed in the public MCP registry as io.github.Khavel/aimodelwatch. Prices, context windows, lifecycle and deprecation dates for 270+ AI/LLM models. Daily, no key. This page covers its npm package (@aimodelwatch/mcp).

### Is the AI Model Watch MCP server safe to use?

AI Model Watch scores 66 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 20 September 2026. It declares no install or post-install scripts. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the AI Model Watch MCP server expose?

AI Model Watch exposes 9 tools: list_models, get_model, search_models, compare_models, cheapest_models, and 4 more. Their descriptions and schemas cost roughly 1,045 tokens of context every time the server is loaded.

### Is the AI Model Watch MCP server still maintained?

AI Model Watch is still listed as active in the MCP registry. We last reached this channel on 20 September 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the AI Model Watch MCP server under?

AI Model Watch declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- npm package: https://www.npmjs.com/package/@aimodelwatch/mcp
- Socket report: https://socket.dev/npm/package/@aimodelwatch/mcp
- Website: https://aimodelwatch.dev/api
- Changelog RSS feed: https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp.json
- HTML version of this page: https://verifymcp.io/servers/khavel-aimodelwatch/aimodelwatch-mcp