Provider pricing
Meta AI API pricing
Meta has 46 priced models in the public snapshot. For 1,000 input and 500 output tokens, Llama 3.2 1b Instruct has the lowest recorded request cost at $0.00003, with $0.02 input and $0.02 output per million tokens.
Best for
Teams that want self-hosting flexibility, open-source transparency, or the lowest possible cost.
Strengths
- Fully open-source — free to self-host with no license fees
- Maverick: 1M context window at $0.20/M input
- Scout: ultra-efficient at $0.10/M input
- Available on every major cloud and inference provider
Tracked Meta models
| Model | Input $/M | Output $/M | Context | Tier | Best for |
|---|---|---|---|---|---|
| Muse Wonder | $4.00 | $20.00 | 1,048,576 | Premium | Meta's top-tier reasoning model |
| Muse Spark Open | $1.25 | $4.25 | 1,048,576 | Mid-tier | Open-weights Muse Spark variant |
| Muse Spark 1.3 | $1.25 | $4.25 | 1,048,576 | Flagship | Meta's latest hosted model with 1M context |
| Llama 4 Maverick | $0.200 | $0.600 | 1,000,000 | Open Source | Free self-host, strong general purpose |
| Llama 4 Scout | $0.100 | $0.250 | 10,000,000 | Open Source | Free self-host, efficient |
| Llama 3 8B | $0.050 | $0.080 | 8,192 | Open Source | Previous-gen 8B open-weight for chat and edge |
| Llama 3.3 70B | $0.200 | $0.600 | 128,000 | Open Source | 70B instruction-tuned open-weight — strong general purpose |
| Llama 3.1 405B | $2.70 | $2.70 | 128,000 | Open Source | Largest open-weight Llama — frontier-class quality |
| Llama 3.1 70B | $0.200 | $0.600 | 128,000 | Open Source | Open-weight 70B for production agents and reasoning |
| Llama 3.1 8B | $0.050 | $0.080 | 128,000 | Open Source | Open-weight 8B for low-latency chat and edge |
| Llama 3 70B | $0.300 | $0.400 | 8,192 | Open Source | Previous-gen Llama 3 70B instruct |
| Llama 4 Behemoth | $0.800 | $2.40 | 256,000 | Premium | 2T-parameter teacher model — frontier open-weight quality |
| Llama 4.1 70B | $0.250 | $0.700 | 1,000,000 | Open Source | Refreshed 70B with 1M context — strong general purpose |
Last updated: · Source: Meta pricing · Methodology
Frequently asked questions
How much does the Meta API cost?
Meta has 46 priced models in the public snapshot. For 1,000 input and 500 output tokens, Llama 3.2 1b Instruct has the lowest recorded request cost at $0.00003, with $0.02 input and $0.02 output per million tokens.
Which Meta model is cheapest?
Llama 3.2 1b Instruct is the lowest-cost recorded model for the 1,000-input/500-output example. Prices are $0.02 input and $0.02 output per million tokens; different workloads may change the ranking.
Does Meta have cached or batch prices?
Cached input: not separately recorded. Batch: not recorded in this dataset. Eligibility and cache-write charges depend on the provider; these discounts are not automatically stacked.
How often are provider prices verified?
The live pricing table synchronizes stored model records daily; public file snapshots update when the site is rebuilt. Synchronization is not independent provider verification. Snapshot updated 2026-10-31; only explicit per-model dates are verification records.