Meta › Llama 4 Maverick 17B 128E Instruct FP8 › Pricing

Llama 4 Maverick 17B 128E Instruct FP8 pricing

Cheapest host: DeepInfra · checked Oct 26, 2026
Benchmarks and alternatives →
$0.20 / $0.80per 1M tokens in / out

Price sheetper 1M tokens

Input$0.20DeepInfra
Output$0.80
Context1Mtokens

Where to buy2 providers · per 1M tokens

ProviderInputOutputContext
DeepInfra$0.20$0.80—
Novita$0.27$0.85—

In Meta's line-upper 1M in / out

Llama 4 Maverick$0.19 / $0.65−13%
Meta Llama 3 70B Instruct$0.30 / $0.40−7%
Llama 3.2 11B Vision Instruct$0.34 / $0.34−1%
Llama 4 Maverick 17B 128E Instruct FP8$0.20 / $0.80this model
Llama 3.2 90B Vision Instruct$0.35 / $0.40+4%
Llama 3 8B$0.30 / $0.60+7%
Llama 3.1 70B Instruct$0.40 / $0.40+14%
<
All Llama API prices →

Cheaper for the same workvs Llama 4 Maverick 17B 128E Instruct FP8

Qwen3 235B A22B Instruct 2507 unstatedOpen-41%score +40%$0.09 / $0.55
GPT-5 Nano (2025-08-07) unstated-61%score +38%$0.05 / $0.40
Qwen3 8B unstatedOpen-11%score +14%$0.18 / $0.70
Qwen3 30B A3B Instruct 2507 unstatedOpen-59%score +11%$0.09 / $0.30
Gemini 2.5 Flash Lite unstated-50%same score$0.10 / $0.40
All alternatives and what breaks when you switch →

Pricing questions

How much does Llama 4 Maverick 17B 128E Instruct FP8 cost?›

$0.20 per 1M input tokens and $0.80 per 1M output tokens.

Is Llama 4 Maverick 17B 128E Instruct FP8 free to use?›

There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.

What is a cheaper alternative to Llama 4 Maverick 17B 128E Instruct FP8?›

Qwen3 235B A22B Instruct 2507 at unstated effort costs -41% per task with a score +40% on GPQA Diamond · Epoch.