Providers /Meta
Inference provider

Meta

Models
5
Data residency
No answer on record
00

Our take

Editorial by LLMap · updated Sep 3, 2026

Meta is a boutique creative-AI provider with a tiny model family built around the Muse Spark 1.x line. Its three tracked offers share an explicit no-training-on-prompts policy, but compliance and geographic details remain unverified in our data.

Use this for budget creative generation when Spark 1-3's quality suffices — it is dramatically cheaper than its siblings. Pick it when prompt privacy matters, since the provider explicitly does not train on prompts. Skip it if you need a deep catalogue, verified compliance attestations, or consistent throughput across tiers.

Strengths
  • Clear price separation between tiers: Spark 1-3 is many times cheaper than Spark 1-1 and 1-2.
  • Explicit no-training-on-prompts policy, stated rather than assumed.
Trade-offs
  • Very shallow catalogue — only 3 tracked offers, all from one model family.
  • Throughput is modest and inconsistent across tiers, with the most expensive variant also the slowest.
  • Compliance and geographic transparency largely absent: headquarters, EU endpoint, SOC 2 and zero-retention option are all unverified in our data.
01

Point your tools here

We have not recorded what a router needs for Meta yet — no base URL and no docs link on file. That is our gap, not a sign Meta has no API; its own documentation is the place to look until we close it.

02

Privacy & data handling

These answers cover requests sent to Meta through OpenRouter, as OpenRouter records them, last read 58 min ago. For Meta's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.

Trains on prompts
✗No
Logs prompts
✓Yes
No answer on record
No answer on record
No answer on record

Prompt retention: 30 days.

03

Models & pricing

Something wrong on this page? Tell us