Skip to content

Meta API pricing

8 Meta models, ranked by what your workload would cost per month. Prices as listed on OpenRouter at 3 Oct 2026, 08:40 UTC.

Your workload

90M input and 18M output tokens a month (30-day month; token counts are per request).

Assumes: 1,000 conversations a day, each sending about 3,000 tokens of history and instructions and receiving about 600 tokens of replies.

8 of 8 models, sorted by your monthly cost — loading the complete list…

Each row: input / output price, then your monthly cost. Token prices per 1M tokens. Tick up to 4 models to compare.

$5.94a month
$6.05a month
$10.44a month
$14.40a month
$14.76a month
$19.44a month
$28.62a month
$43.20a month

“Your cost” applies the cache-read price only where the source lists one; a model with no price data shows a dash and is never ranked as cheapest.

Collected from the OpenRouter models API (466 entries). 99 models are cross-checked against provider list prices in the LiteLLM price table.

Meta prices at a glance

  • 8 models listed, 8 of them with a paid price. Batch variants are folded into their base model.
  • Lowest input price: Llama 3.2 1B Instruct, at $0.027 per million input tokens and $0.201 per million output tokens.
  • Highest input price: Llama 3.1 70B Instruct, at $0.40 input and $0.40 output per million tokens.
  • 1 models publish a cache-read price; the median discount on input is 50% (from 50% to 50%).

These are OpenRouter prices. Open a model in the table to see the provider list price recorded by LiteLLM, where there is one, and whether the two differ. To understand the discounts, read the guide to prompt caching and batch pricing.

Meta models with their own page

Questions and answers

How many Meta models are listed?

8 at the collection of 3 Oct 2026, 08:40 UTC, of which 8 have a paid price on OpenRouter. Batch variants are folded into their base model.

What is the cheapest Meta model?

By input price, Llama 3.2 1B Instruct: $0.027 per million input tokens and $0.201 per million output tokens. Whether it is the cheapest for you depends on your mix of input and output, which is what the table on this page calculates.

What is the most expensive Meta model?

By input price, Llama 3.1 70B Instruct: $0.40 per million input tokens and $0.40 per million output tokens.

Do Meta models have cached-input prices?

1 of the 8 listed models publish a cache-read price. The table shows it in the “Cached in” column.

Are these the prices Meta charges directly?

They are the prices on OpenRouter. Where LiteLLM records the provider list price for a model, opening the model shows both, and the row is flagged when they differ by more than 1%.