Llama 4 Maverick API Prices Rise 33%: What It Costs You Now

The Price Change

Meta's Llama 4 Maverick has received its first notable price increase. Input tokens now cost $0.20 per million (up from $0.15), and output tokens move to $0.80 per million (up from $0.60) — a clean +33% across the board.

The ratio between input and output pricing stays the same (1:4), so the structure hasn't changed, just the baseline.

Does This Actually Hurt?

It depends heavily on your workload mix.

• Low-volume or input-heavy workloads (retrieval, classification, short completions): the jump from $0.15 → $0.20 on input is small in absolute terms. A pipeline consuming 10M input tokens monthly goes from $1.50 to $2.00.
• High-volume, output-heavy workloads (drafting, summarization, agents): the $0.60 → $0.80 output increase hits harder. 50M output tokens monthly jumps from $30 to $40 — an extra $120/year at that scale, more if you're running larger volumes.

33% is a meaningful jump in a market where competing open-weight models have been trending flat or downward on price. If Maverick's quality justified the old price for your use case, run the math again before assuming it still does.

What to Do Next

If you're feeling the squeeze, it's worth benchmarking Maverick against current alternatives on quality and cost per task — not just sticker price. A model that's 20% cheaper but requires more output tokens to complete a task can end up costing more.

Check the Llama 4 Maverick — live specs & price history page for up-to-date pricing and side-by-side comparisons with other models in the same capability tier.

Ezra, Scout AI Team