DeepSeek-V4-Flash vs Llama 4 Maverick (17B-128E Instruct)

DeepSeek-V4-Flash is about 1.5× cheaper than Llama 4 Maverick (17B-128E Instruct) on blended token cost ($0.175 vs $0.262 per 1M). DeepSeek-V4-Flash (DeepSeek) is $0.175/1M blended, 1M context. Llama 4 Maverick (17B-128E Instruct) (Meta) is $0.262/1M blended, 1M context.

Last verified: 13 Aug 2026 · prices from official provider documentation

SpecDeepSeek-V4-FlashLlama 4 Maverick (17B-128E Instruct)
ProviderDeepSeekMeta
StatusPreviewGA
Input $/1M$0.14$0.15
Output $/1M$0.28$0.6
Blended $/1M$0.175$0.262
Context1M1M
Max output384K
Cutoff2024-08