Meta API Pricing (2026)
Official token rates, aggregator discounts, and monthly cost simulations for Meta.
17 models tracked
META PRICING SUMMARY
How much does Meta API cost?
How much does Meta API cost?
Meta model rates start at $0.020 per 1M input tokens (Llama 3.1 8B Instruct) and up to $0.34 per 1M output tokens for flagship Llama 4 Scout 17B 16E Instruct. Prompt Caching reduces input costs down to $0.06/1M on cache hits.
AI Token Cost Calculator
Simulate your monthly API expenses with Prompt Caching discounts & Aggregator deals.
Input Tokens (Prompt)1,500
10032k128k
Output Tokens (Completion)800
504k16k
Prompt Cache Hit30%
0% (No Cache)50%100% (Hot)
Daily API Calls (Volume Scale):500 calls / day (15,000 / mo)
Single Call Cost
$0.00041
Per request average cost
Official Monthly Bill
$6.18 / mo
Direct Lab rate: $0.11/1M in · $0.34/1M out
Via Hosting & Relays
$5.55 / mo
Saves $0.64/mo (10.3%)
Meta Models Price Comparison Table
Tier:
| Cached In | Hosting Providers & Relays | Action | ||||
|---|---|---|---|---|---|---|
Llama 3.1 8B InstructBest Value | 128k | $0.020 | $0.0100 | $0.05 | $0.02 / $0.05via Cerebras Cerebras (1800 t/s)Groq (560 t/s)DeepInfra | |
Llama 3.1 8B InstructBest Value | 128k | $0.020 | $0.0100 | $0.05 | $0.02 / $0.05via Novita AI Novita AIOpenRouter | |
Llama 3.1 8B InstantBest Value | 128k | $0.050 | $0.0250 | $0.08 | $0.05 / $0.08via Groq Groq (390 t/s)OpenRouter | |
| 1.0M | $0.10 | $0.0020 | $0.20 | $0.10 / $0.20via OpenRouter OpenRouter | ||
| 1.0M | $0.10 | $0.0020 | $0.20 | $0.10 / $0.20via OpenRouter OpenRouter | ||
| 128k | $0.11 | $0.06 | $0.34 | $0.10 / $0.30via Together AI Save $Groq (390 t/s)Together AIOpenRouter | ||
Llama 3.3 70B InstructFlagship | 128k | $0.13 | $0.07 | $0.40 | $0.13 / $0.40via Cerebras Cerebras (1800 t/s)Groq (300 t/s)DeepInfra (80 t/s)Together AIOpenRouter | |
| 128k | $0.14 | $0.07 | $0.40 | $0.14 / $0.40via Novita AI Novita AIOpenRouter | ||
Llama 3 8B Instruct LiteBest Value | 128k | $0.14 | $0.07 | $0.14 | $0.14 / $0.14via Together AI Together AIOpenRouter | |
| 128k | $0.15 | $0.07 | $0.60 | $0.15 / $0.60via Together AI Together AINovita AIOpenRouter | ||
Llama 4 Scout InstructFlagship | 128k | $0.18 | $0.09 | $0.59 | $0.18 / $0.59via Novita AI Novita AIOpenRouter | |
Llama 3.3 70B VersatileFlagship | 128k | $0.59 | $0.29 | $0.79 | $0.59 / $0.79via Groq Groq (390 t/s)Together AIOpenRouter | |
| 128k | $0.80 | $0.40 | $2.00 | $0.80 / $2.00via Together AI Together AIDeepInfraOpenRouter | ||
| 1.0M | $1.25 | $0.15 | $4.25 | $0.10 / $0.20via OpenRouter Save $OpenRouter | ||
| 1.0M | $1.25 | $0.15 | $4.25 | $1.25 / $4.25via OpenRouter OpenRouter | ||
| 1.0M | $1.25 | $0.15 | $4.25 | $0.10 / $0.20via OpenRouter Save $OpenRouter | ||
Llama 3.1 405BFlagship | 128k | $3.50 | $1.75 | $3.50 | $1.00 / $1.00via OpenRouter Save $Together AIOpenRouter |
Price Indicator: Budget (<$0.15) Mid ($0.15 - $1.00) Higher ($1.00 - $5.00) Flagship (>$5.00)
Showing 17 of 17 AI models
Frequently Asked Questions
Everything you need to know about AI token costs, caching discounts, and billing.
How much does Meta API cost per token?
Meta charges on a per-million token basis. For Llama 4 Scout 17B 16E Instruct, input tokens cost $0.11/1M and output tokens cost $0.34/1M.
Does Meta offer Prompt Caching discounts?
Yes. Meta supports automated or explicit prompt caching, reducing input costs on cache hits down to $0.055/1M.
Can I access Meta models via OpenRouter or other aggregators?
Yes, most Meta models are mirrored or directly routed through aggregators like OpenRouter, DeepInfra, and Together AI, often offering pay-as-you-go billing without minimum monthly commitments.