AI & agents · Tool integration
meta-model-api-quickstart
/meta-model-api-quickstart
Use when making your first Meta Model API call.
For whom
- Author
- NullAITech ↗
- Repository
- meta-dev-skills ↗
- Licence
- MIT
- Last repository change
- 30.10.2026
Assessment and security
96/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability21 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
Deprecation
No signals detected. Repository changed on 30.10.2026, without references to retired models or APIs.
Licence
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Use when making your first Meta Model API call.
Original author description
Use when making your first Meta Model API call.
How it works
- When to Use
- 1. Confirm the account prerequisites
- 2. Get the real identifiers from the docs
- 3. Make a first call
- 4. Verify the call
- Pitfalls
- Done when
Original SKILL.md headings · 88 lines, 591 words
SKILL.md preview
---
name: meta-model-api-quickstart
description: "Use when making your first Meta Model API call."
version: 0.1.0
author: "Zoth / NullAI"
license: "MIT"
metadata:
hermes:
tags: [meta, model-api, quickstart, credits, auth]
related_skills: [muse-spark-multimodal-agents, sam3-segmentation, meta-open-weight-models]
---
# Meta Model API quickstart
Get a Meta Model API call working end to end: account, credits, auth, choosing the
right model, a first request, and a spend guard.
## When to Use
- You are about to call a Meta Model API model for the first time.
- You need to set up auth, API keys, or base URLs for Meta's hosted models.
- You need to claim, track, or limit API credits and spend.
- You need the **correct, callable** model ID for a Meta Model API model.
- You keep getting 401 / 403 / 404 or "model not found" errors.
Do not use this for self-hosted serving of Meta's open weights — see
`meta-open-weight-models` instead.
## 1. Confirm the account prerequisites
- Credits must be claimed for the account. Each accepted participant claims their
own allocation; it is non-transferable. Claim before the deadline.
- **A credit card must be on the account** — the API is not callable without it.
- Add a **spend-limit notification** immediately after the card is added, before
the first call. Build loops later, not before the limit exists.
## 2. Get the real identifiers from the docs
Never invent the base URL, auth header, request body, or model ID. Read, in order:
1. **Overview** — what the API is and which models it serves.
2. **Quickstart** — base URL, authentication, and a minimal request.
3. **Model list** — the exact, callable model IDs.
4. **API cookbook** — request and response shapes per model.
Copy exact strings from those pages. A model ID on a marketing table is not
always the callable ID.
## 3. Make a first call
Keep credentials and identifiers in the environment, never in source:
```sh
export META_API_BASE_URL="<base URL from the Quickstart>"
export META_API_KEY="<your key>"
export META_MODEL_ID="<exact id from the Model list>"
```
Send the minimal request exactly as the Quickstart and Model list specify for the
model you picked. The request **path, auth header, and body field names are
model- and version-specific** — the ones below are placeholders:
```sh
curl -sS "$META_API_BASE_URL/<path from the Model list>" \
-H "Authorization: Bearer $META_API_KEY" \
-H "Content-Type: application/json" \
--data "$(cat request.json)"
```
Replace `<path>`, `Authorization`, and the keys in `request.json` with the values
shown in the Model list / Quickstart for that model. Do not copy a path or field
name from a blog post or another model's example.
## 4. Verify the call
- Success is a **2xx response with a non-empty body** — not merely "the command
exited 0".
- Log the response's request ID and any usage/cost fields; you need them for
support and to attribute spend against the credit budget.
- On a model-not-found error, re-read the Model list: the ID is almost always the
problem, not the endpoint.
## Pitfalls
- **Guessing model IDs.** IDs differ between a docs page and a partner/hackathon
deployment. The same open weights can be exposed under a *different, suffixed*
ID by a hosting partner. Take the ID from the surface you are actually calling.
- **No card, no call.** Credits alone are insufficient; the API returns auth
errors without a payment method on the account.
- **No spend limit.** A runaway retry loop burns the allocation fast. Set the
limit before you build the loop.
- **Secrets in git.** Projects are often public. Keep keys in environment
variables or a git-ignored file — never commit them.
## Done when
A minimal request returns 2xx with a non-empty body, the request ID is captured,
and a spend limit/notification is configured.
Versions
The database has no other skills named meta-model-api-quickstart.
Database snapshot: 13.11.2026 · package version 4cfad6d4af3b (SHA-256 digest of the entire package)
Downloads in the last 30 days: 1 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100meta-reasoning
Use for text reasoning, reading photos, understanding video, reading PDFs, or getting JSON back from Muse Wonder through the Meta Model API Responses endpoint. - 90onboard-quickstart
Write a developer quickstart — minimal steps from zero to first successful API call. Use when asked to "write a quickstart", "get users from zero to first API call", or "write a getting started guide". - 100quickstart-guide-generator
Generate quickstart guide generator operations. Auto-activating skill for Technical Documentation. Triggers on: quickstart guide generator, quickstart guide generator Part of the Technical Documentation skill category. Use when working with quickstart guide generator… - 100meta-mcp-builder
Build Model Context Protocol (MCP) servers that expose services, APIs, and data as typed tools for agents. Use when "build an MCP server", "give Claude access to X", "create an MCP tool", "expose my API to an agent", or "AI agent integration". - 100langchain-python-quickstart
Scaffold a minimal local LangChain agent in Python by following the official quickstart. Use when the user wants to quickly build or try a LangChain agent locally. - 100langchain-typescript-quickstart
Scaffold a minimal local LangChain agent in TypeScript by following the official quickstart. Use when the user wants to quickly build or try a LangChain agent locally.