Meta API Pricing
Compare input and output token prices across AI language models.
| Provider | Model | Input per 1M tokens | Output per 1M tokens | |
|---|---|---|---|---|
Meta | Llama 3.1 8BOpen weights | $0.02 | $0.05 | View |
Meta | Llama 3.2 1BOpen weights | $0.03 | $0.20 | View |
Meta | Llama 3 8BOpen weights | $0.05 | $0.25 | View |
Meta | Llama 3.2 11B VisionOpen weights | $0.05 | $0.68 | View |
Meta | Llama 3.2 3BOpen weights | $0.05 | $0.34 | View |
Meta | Muse Ember | $0.10 | $0.50 | View |
Meta | Llama 3.3 70BOpen weights | $0.14 | $0.40 | View |
Meta | Llama 4 ScoutOpen weights | $0.16 | $0.64 | View |
Meta | Muse Glimmer 30BOpen weights | $0.20 | $0.80 | View |
Meta | Llama 4 MaverickOpen weights | $0.25 | $0.87 | View |
Meta | Llama 3 70BOpen weights | $0.65 | $2.75 | View |
Meta | Muse Spark 1.1 | $1.25 | $4.25 | View |
Meta | Muse Spark 1.2 | $1.25 | $4.25 | View |
Meta | Muse Spark 1.3 | $1.25 | $4.25 | View |
Meta | Muse Spark OpenOpen weights | $1.25 | $4.25 | View |
Meta | Muse Wonder | $4.00 | $20.00 | View |
Meta API cost calculator
Enter the tokens of a typical request to compare what it costs with every model in the table. Not sure how many tokens your prompt has? Paste it into the token calculator.
| Model | Per request | Per month |
|---|---|---|
| Llama 3.1 8B (open weights, third-party host) | $0.000065 | $0.65 |
| Llama 3.2 1B (open weights, third-party host) | $0.00015 | $1.55 |
| Llama 3 8B (open weights, third-party host) | $0.00022 | $2.25 |
| Llama 3.2 3B (open weights, third-party host) | $0.00027 | $2.70 |
| Llama 3.2 11B Vision (open weights, third-party host) | $0.00044 | $4.36 |
| Muse Ember | $0.00045 | $4.50 |
| Llama 3.3 70B (open weights, third-party host) | $0.00047 | $4.70 |
| Llama 4 Scout (open weights, third-party host) | $0.00064 | $6.40 |
| Muse Glimmer 30B (open weights, third-party host) | $0.0008 | $8.00 |
| Llama 4 Maverick (open weights, third-party host) | $0.00093 | $9.35 |
How Meta API pricing works
Meta bills its API per token, with separate prices for input tokens (your prompt) and output tokens (the generated answer). The table shows standard rates per 1 million tokens; a token is roughly three quarters of an English word.
FAQ
How much does the Meta API cost?
As of November 16, 2026, Meta charges from $0.10 to $4.00 per 1M input tokens and from $0.50 to $20.00 per 1M output tokens across the 5 models in the table above. Meta also publishes 12 open-weight models that you run on your own hardware or through third-party hosts, so their price depends on the host.
What is the cheapest Meta model?
Muse Ember, at $0.10 per 1M input tokens and $0.50 per 1M output tokens. A request with 2,000 input and 500 output tokens costs about $0.00045 with it, so 1,000 such requests cost $0.45.
What is the most expensive Meta model?
Muse Wonder, at $4.00 per 1M input tokens and $20.00 per 1M output tokens. The same 2,000 input and 500 output token request costs about $0.02, so pick the smallest model that handles your task well.
Which Meta model has the largest context window?
Muse Ember accepts up to 1M tokens of context. Long prompts are billed as input tokens, so filling the whole window costs about $0.10 per request.
Is Meta cheaper than other LLM APIs?
Not at the low end: Qwen3.7-Flash costs $0.03 per 1M input tokens and $0.13 per 1M output tokens, less than any Meta model. Price is only part of the choice, so compare quality and context window as well in the LLM price comparison.
How do I estimate the cost of my prompts?
Count the tokens of a typical prompt and answer with the token calculator, then enter them in the cost calculator above to see the price per request and per month for every Meta model.
