Llama 4 Maverick 17B 128E Instruct FP8 API pricing
Updated
Among the 7 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest blended standard price for Llama 4 Maverick 17B 128E Instruct FP8 is GMI Cloud at USD 0.25 per 1M input tokens and USD 0.80 per 1M output tokens. The median across 7 standard prices is USD 0.27 / USD 1. Meta sells no API of its own for it.
Does Meta sell Llama 4 Maverick 17B 128E Instruct FP8 directly?
No. Meta does not sell Llama models through an API of its own: llama.com now redirects to Meta's developer site, whose Model API lists only Meta's Muse models. Llama is available only from third-party hosts. Checked 2026-11-14.
Llama 4 Maverick 17B 128E Instruct FP8 model facts
Facts from Meta's model card for meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 and the Hugging Face model API, read 2026-11-16. BroadNet did not verify them, and hosts may serve the model at a different precision. BroadRouter is BroadNet's site.
| Fact | As stated | Source |
|---|---|---|
| Open weights | Yes, with gated access. meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8 is public, but Hugging Face asks you to request access and accept the terms before you download (gated: manual). | Hugging Face model API, checked 2026-11-16Quote“"private": false, "gated": "manual", 84 .safetensors files listed” |
| Licence | Llama 4 Community License Agreement (custom commercial license) | Meta model card, checked 2026-11-16Quote“A custom commercial license, the Llama 4 Community License Agreement” |
| Total parameters | 400B | Meta model card, checked 2026-11-16Quote“Llama 4 Maverick (17Bx128E) | 17B (Activated) 400B (Total)” |
| Active parameters per token | 17B | Meta model card, checked 2026-11-16Quote“Llama 4 Maverick (17Bx128E) | 17B (Activated) 400B (Total)” |
| Context length | 1M | Meta model card, checked 2026-11-16Quote“Multilingual text and code | 1M” |
| Input and output | Multilingual text and image in; multilingual text and code out | Meta model card, checked 2026-11-16Quote“Multilingual text and image | Multilingual text and code” |
| Tool calling | Not found on Meta's model card when we read it | checked 2026-11-16 |
| Release date | 2025-04-05 | Meta model card, checked 2026-11-16Quote“Model Release Date: April 5, 2025” |
| First published on Hugging Face (not necessarily the release date) | 2025-04-01 | Hugging Face model API, checked 2026-11-16Quote“"createdAt": "2025-04-01T20:47:02.000Z"” |
Llama 4 Maverick 17B 128E Instruct FP8 price per provider
USD per 1M tokens. Standard prices are ranked on a blend of 3 input tokens to 1 output token, cheapest first; the dearest standard price is 68% above the cheapest. Free rows, the maker's serving version and other service tiers follow, tagged and unranked.
| # | Provider | Input | Cached input | Output | Blended 3:1 | Context | Precision | Source |
|---|---|---|---|---|---|---|---|---|
| 1 | GMI Cloud cheapestmeta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8Notes and source quoteNo cache-read price published. “Price (per 1M tokens) Input $0.25 Output $0.80” | USD 0.25 | not published | USD 0.80 | USD 0.3875 | 1,048,576 | FP8 | GMI Cloud page checked 2026-11-15 re-checked 2026-11-16 |
| 2 | Novita AImeta-llama/llama-4-maverick-17b-128e-instruct-fp8Notes and source quote“Llama 4 Maverick Instruct 1M $0.27 /Mt $0.85 /Mt” | USD 0.27 | not published | USD 0.85 | USD 0.415 | 1,048,576 | not stated | Novita AI page checked 2026-11-14 re-checked 2026-11-16 |
| 3 | Amazon BedrockNotes and source quoteShown for region US East (N. Virginia) on the pricing page (values filled from AWS's published price list for us-east-1). “Llama 4 Maverick 17B $0.24 $0.97 $0.12 $0.485” | USD 0.24 | not published | USD 0.97 | USD 0.4225 | not stated | not stated | Amazon Bedrock page checked 2026-11-14 re-checked 2026-11-16 |
| 4 | Microsoft AzureNotes and source quoteFoundry Models serverless (pay-as-you-go), Global deployment. “Llama4 Mavrick 17B Global $0.25 $1” | USD 0.25 | not published | USD 1 | USD 0.4375 | not stated | not stated | Microsoft Azure page checked 2026-11-14 re-checked 2026-11-16 |
| 5 | ParasailNotes and source quote“Llama 4 Maverick (FP8) $0.35 $1.00 $0.17” | USD 0.35 | USD 0.17 | USD 1 | USD 0.5125 | not stated | FP8 | Parasail page checked 2026-11-14 re-checked 2026-11-16 |
| 6 | Google Vertex AIllama-4-maverickNotes and source quote“Llama 4 Maverick Input $0.35 … Output $1.15” | USD 0.35 | not published | USD 1.15 | USD 0.55 | not stated | not stated | Google Vertex AI page checked 2026-11-14 re-checked 2026-11-16 |
| 7 | IBM watsonx.aillama-4-maverick-17b-128e-instruct-fp8Notes and source quote“llama-4-maverick-17b-128e-instruct-fp8 Meta USD 0.371 per 1M tokens input USD 1.484 per 1M tokens output” | USD 0.371 | not published | USD 1.484 | USD 0.6492 | not stated | FP8 | IBM watsonx.ai page checked 2026-11-15 re-checked 2026-11-16 |
| — | Amazon Bedrock batchNotes and source quoteBatch columns. Shown for region US East (N. Virginia) on the pricing page (values filled from AWS's published price list for us-east-1). “Llama 4 Maverick 17B $0.24 $0.97 $0.12 $0.485” | USD 0.12 | not published | USD 0.485 | USD 0.2112 | not stated | not stated | Amazon Bedrock page checked 2026-11-14 re-checked 2026-11-16 |
| — | Google Vertex AI batchllama-4-maverickNotes and source quote“Llama 4 Maverick Batch Input $0.175 … Batch Output $0.575” | USD 0.175 | not published | USD 0.575 | USD 0.275 | not stated | not stated | Google Vertex AI page checked 2026-11-14 re-checked 2026-11-16 |
- Ranked: standard-tier prices only, on (3 × input + output) ÷ 4. "Cached input: not published" means the provider's page shows no cache price; budget cached tokens at the input price.
- Other tiers (batch): batch runs asynchronously within hours; flex and spot are best-effort and can be slow or refused at capacity; priority and fast cost more for reliability or speed. Shown, not ranked.
- Automated re-check 2026-11-16: a weekly job re-fetches each source page and looks for the recorded quote or prices. It is not a human review. "Re-checked" means it still found them. Rows with no re-check note show only the date the price was read.
- Precision is shown only where the provider states it; "not stated" is unknown, not full precision.
Report a wrong price or number (email to info@broadnet.ai, with this page's link filled in).
What does Llama 4 Maverick 17B 128E Instruct FP8 cost for a real workload?
Monthly cost over 30 days at each standard price, using the presets of the LLM API cost calculator. Providers without a cached price bill cached input at the input price.
- Support chatbot: 2,000 requests a day, 1,500 input tokens (60% cached), 300 output.
- RAG over documents: 1,000 requests a day, 8,000 input tokens (30% cached), 500 output.
- Coding agent (one developer): 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning.
| Provider | Support chatbot | RAG over documents | Coding agent (one developer) |
|---|---|---|---|
| GMI Cloud | USD 36.90 | USD 72.00 | USD 522.00 |
| Novita AI | USD 39.60 | USD 77.55 | USD 562.50 |
| Amazon Bedrock | USD 39.06 | USD 72.15 | USD 519.30 |
| Microsoft Azure | USD 40.50 | USD 75.00 | USD 540.00 |
| Parasail | USD 39.78 | USD 86.04 | USD 444.60 |
| Google Vertex AI | USD 52.20 | USD 101.25 | USD 733.50 |
| IBM watsonx.ai | USD 60.10 | USD 111.30 | USD 801.36 |
For the support chatbot, the cheapest price (GMI Cloud, USD 36.90 a month) saves USD 2.88 against the median (USD 39.78). For the coding agent the gap is USD 95.40 a month (18% lower). The winner changes with the workload: Parasail wins the cache-heavy agent, GMI Cloud the chatbot.
Which Llama 4 Maverick 17B 128E Instruct FP8 provider is cheapest for your traffic?
Amazon Bedrock has the lowest input price (USD 0.24) and GMI Cloud the lowest output price (USD 0.80). They cost the same at 17 input tokens per output token. With more input than that, as with long prompts or retrieval, Amazon Bedrock is cheaper; with less, as with long answers, GMI Cloud is.
| Provider | No cache hits | 50% cached | 90% cached |
|---|---|---|---|
| GMI Cloud (no cache price) | USD 0.3875 cheapest | USD 0.3875 cheapest | USD 0.3875 cheapest |
| Novita AI (no cache price) | USD 0.415 | USD 0.415 | USD 0.415 |
| Amazon Bedrock (no cache price) | USD 0.4225 | USD 0.4225 | USD 0.4225 |
| Microsoft Azure (no cache price) | USD 0.4375 | USD 0.4375 | USD 0.4375 |
| Parasail | USD 0.5125 | USD 0.445 | USD 0.391 |
| Google Vertex AI (no cache price) | USD 0.55 | USD 0.55 | USD 0.55 |
USD per 1M tokens, 3 input tokens to 1 output token, each provider's cheapest standard row; 1 dearer providers are in the price table above. A provider without a cached price bills cached input at the input price. Any charge for writing to the cache, which some providers add, is not included.
1 of 7 providers publishes a cached-input price below the input price: Parasail (51% off). GMI Cloud, Novita AI, Amazon Bedrock, Microsoft Azure, Google Vertex AI and IBM watsonx.ai publish none, so cached tokens there bill at the input price.
GMI Cloud and Novita AI state a context window of 1,048,576 tokens; Amazon Bedrock, Microsoft Azure, Parasail, Google Vertex AI and IBM watsonx.ai do not state one.
What does Llama 4 Maverick 17B 128E Instruct FP8 cost in other languages?
The same text needs a different number of tokens in each language. With the Llama 4 Maverick tokenizer, sending 1M characters costs this much at the cheapest standard input price we found for Llama 4 Maverick 17B 128E Instruct FP8:
| Content | Input per 1M characters | Cheapest provider |
|---|---|---|
| English | USD 0.051 | Amazon Bedrock |
| Arabic MSA | USD 0.089 | Amazon Bedrock |
| Hindi | USD 0.082 | Amazon Bedrock |
| Chinese | USD 0.159 | Amazon Bedrock |
| Code | USD 0.061 | Amazon Bedrock |
From BroadRouter's real cost by language dataset, which covers 15 languages and content types for every model we could match to a measured tokenizer.
Estimate your own bill
- LLM API cost calculator with Llama 4 Maverick 17B 128E Instruct FP8 preloaded: your traffic and cache rate, against any of 428 models.
- Llama 4 token counter: count the tokens in your real prompts. Tokenizers differ, so the same text costs a different number of tokens on different models.
- Prompt caching calculator: what a higher cache-hit rate saves.
How does Llama 4 Maverick 17B 128E Instruct FP8 compare with other Meta models?
The Meta models whose cheapest standard price is closest to Llama 4 Maverick 17B 128E Instruct FP8's (USD 0.3875 blended):
| Model | Cheapest input | Cheapest output | Max context | Providers |
|---|---|---|---|---|
| Muse Glimmer 30B | USD 0.30 | USD 1.10 | 131,072 | 3 |
| Llama 3.1 Instruct (70B) | USD 0.72 | USD 0.72 | 32,768 | 3 |
| Llama 3.2 11B Vision Instruct | USD 0.16 | USD 0.16 | — | 3 |
| Llama 3.3 70B | USD 0.10 | USD 0.32 | 131,072 | 14 |
| Llama 4 Scout 17B 16E Instruct | USD 0.10 | USD 0.30 | 1,048,576 | 6 |
| Llama 3.2 3B Instruct | USD 0.0509 | USD 0.335 | 128,000 | 4 |
Priced closest from other makers: Qwen3 235B A22B FP8 (Qwen, USD 0.40) · DeepSeek R1 Distill Llama 70B (DeepSeek, USD 0.375) · Qwen-MT Plus (Qwen, USD 0.375).
Compare Llama 4 Maverick 17B 128E Instruct FP8 with
- Llama 4 Maverick vs Llama 3.3 70B: API price side by side
- Llama 4 Maverick vs DeepSeek V4.1 Flash: API price side by side
- Qwen3.8 27B vs Llama 4 Maverick: API price side by side
Also compared: vs Llama 3.3 70B · vs DeepSeek V4.1 Flash · vs Qwen3.8 27B.
Every Llama 4 Maverick 17B 128E Instruct FP8 price comparison · All model-vs-model price comparisons
All 428 models: API prices compared · Cheapest LLM APIs right now · LLM API price index (open data) · Download all prices (CSV, JSON) · Price change log
Questions
How much does the Llama 4 Maverick 17B 128E Instruct FP8 API cost per token?
From USD 0.24 per 1M input tokens (USD 0.00000024 per token) and USD 0.80 per 1M output tokens (USD 0.0000008 per token) at the lowest standard prices we found, which can come from different providers. The median of 7 standard prices is USD 0.27 input and USD 1 output. Each price was read from the provider's own pricing page; the source is linked in its row.
Which provider is cheapest for Llama 4 Maverick 17B 128E Instruct FP8?
Among the 7 providers we checked (prices read 2026-11-14 to 2026-11-15), the lowest is GMI Cloud, ranked on a blend of 3 input tokens to 1 output token (USD 0.3875 per 1M). For input alone the lowest price is USD 0.24 (Amazon Bedrock); for output alone, USD 0.80 (GMI Cloud). Batch, flex, priority and fast tiers are shown in the table but not ranked.
Does Meta sell Llama 4 Maverick 17B 128E Instruct FP8 on its own API?
Meta does not sell Llama models through an API of its own: llama.com now redirects to Meta's developer site, whose Model API lists only Meta's Muse models. Llama is available only from third-party hosts. The lowest-priced host we checked is GMI Cloud at USD 0.25 / USD 0.80 per 1M tokens.
Does Llama 4 Maverick 17B 128E Instruct FP8 have cached input pricing?
1 of 7 standard prices includes one. The lowest is USD 0.17 per 1M at Parasail. Providers without one bill cached input at the input price. What cached input means.
What is the Llama 4 Maverick 17B 128E Instruct FP8 context window?
1,048,576 tokens on every provider that states it.
How much does a coding agent cost per month on Llama 4 Maverick 17B 128E Instruct FP8?
For 1,500 requests a day, 40,000 input tokens (85% cached), 2,000 output incl. reasoning, over 30 days: USD 444.60 a month at Parasail, USD 540.00 at the median price and USD 801.36 at IBM watsonx.ai. Try your own numbers.
Is there a batch, flex or priority price for Llama 4 Maverick 17B 128E Instruct FP8?
Yes: batch from Amazon Bedrock at USD 0.12 / USD 0.485 per 1M input / output tokens. These tiers trade speed or availability for price (batch runs within hours, flex and spot can be slow or refused at capacity, priority and fast cost more for speed or reliability), so they are shown but not ranked.
Is Llama 4 Maverick 17B 128E Instruct FP8 free to use through an API?
No provider in our data lists a free price for Llama 4 Maverick 17B 128E Instruct FP8 as of 2026-11-15. Free tiers and credits by provider (checked 2026-11-14) lists general offers that do not name a model; all free LLM API models are on one page.
Is Llama 4 Maverick 17B 128E Instruct FP8 open source, with open weights, and what is its licence?
Meta publishes the weights on Hugging Face as meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8, with gated access: Hugging Face asks you to request access and accept the terms before you download. The licence, as the Meta model card states it: Llama 4 Community License Agreement (custom commercial license). Open weights means you can download and run the model yourself; what you may do with it, including commercial use, is set by that licence, so read it in full (checked 2026-11-16). These facts come from Meta's card and the Hugging Face model API; BroadNet did not verify them, and hosts may serve the model at a different precision.
How many parameters does Llama 4 Maverick 17B 128E Instruct FP8 have?
Total parameters, as the Meta model card states them: 400B. Active parameters per token: 17B. (checked 2026-11-16; BroadNet did not verify these figures.)
Read as MarkdownAsk ChatGPTAsk ClaudeAsk Perplexity
Sources
- Amazon Bedrock: pricing (checked 2026-11-14)
- Microsoft Azure: pricing (checked 2026-11-14)
- GMI Cloud: pricing (checked 2026-11-15)
- Google Vertex AI: pricing (checked 2026-11-14)
- IBM watsonx.ai: pricing (checked 2026-11-15)
- Novita AI: pricing (checked 2026-11-14)
- Parasail: pricing (checked 2026-11-14)
- Meta developer docs (checked 2026-11-14)