Llama 4 Maverick API price — 2 gateways compared
Cheapest right now: OpenRouter at $0.19/MTok in
Meta · from $0.19 per
million input tokens · 1.1× spread between gateways
· updated continually
Gateway $ in /MTok $ out /MTok
Context Input price history Model ID OpenRouter fees $0.19 $0.65 1,048,576 meta-llama/llama-4-maverick use Glama $0.20 $0.80 1,048,576 meta/llama-4-maverick use
Fees beyond the token price OpenRouter : No markup on token prices (provider pass-through), but credit purchases cost 5.5% via card ($0.80 minimum) or 5% via crypto; BYOK usage above the monthly allowance is charged 5%. source Changelog for Llama 4 Maverick
2026-09-22 openrouter cut meta-llama-llama-4-maverick-output 18.4%: 0.8 → 0.6525 usd_per_mtok 2026-09-22 openrouter cut meta-llama-llama-4-maverick-input 6.3%: 0.2 → 0.1875 usd_per_mtok 2026-09-21 openrouter raised meta-llama-llama-4-maverick-output 22.6%: 0.6525 → 0.8 usd_per_mtok 2026-09-21 openrouter raised meta-llama-llama-4-maverick-input 6.7%: 0.1875 → 0.2 usd_per_mtok 2026-09-14 openrouter cut meta-llama-llama-4-maverick-output 6.2%: 0.696 → 0.6525 usd_per_mtok 2026-09-14 openrouter cut meta-llama-llama-4-maverick-input 6.3%: 0.2 → 0.1875 usd_per_mtok 2026-08-31 openrouter cut meta-llama-llama-4-maverick-output 13.0%: 0.8 → 0.696 usd_per_mtok 2026-08-27 requesty raised llama-4-maverick-17b-128e-instruct-fp8-output 11.1%: 0.765 → 0.85 usd_per_mtok 2026-08-27 requesty raised llama-4-maverick-17b-128e-instruct-fp8-input 11.1%: 0.18 → 0.2 usd_per_mtok Llama 4 Maverick pricing FAQ
How much does Llama 4 Maverick cost per million tokens? The cheapest tracked gateway charges $0.19 per million input tokens and $0.65 per million output tokens (OpenRouter). Prices refresh continually.
Which gateway is cheapest for Llama 4 Maverick? OpenRouter for input tokens and OpenRouter for output tokens, among the 2 gateways we track.
Do Llama 4 Maverick prices differ between gateways? Yes — the most expensive tracked gateway charges 1.1× the cheapest for input tokens. Identical model, identical tokens, different bill.
← All LLM prices · API
Every datum links its source and is versioned in
public git history .
Data: CC BY 4.0 — cite gpudiff.com. Nameplate specs are vendor claims, not measured
throughput. Vast prices are the 25th percentile of verified marketplace listings
(what a careful buyer actually gets); RunPod prices are list. Outbound provider
links may carry referral codes and sponsored units are labelled; neither ever
affects the numbers. Sponsor this site .