LLM PricesModelsChangelog
Prices in USD per 1M tokens, text only. Each price shows where it came from and when it was last verified. Prices marked unverified come from models.dev (community-maintained) until we read them from the provider directly.
← All models

DeepSeek V4 Flash

Same open weights, served by 1 provider. Weights: deepseek-ai/DeepSeek-V4-Flash.

deepseekOpen weights1M contextReleased 2026-04-24deepseek/deepseek-v4-flash

Providers

Standard tier onlyFull precision only
ProviderTheir model idServingInputOutputCached inBlended 3:1Source
DeepInfra
deepseek-ai/DeepSeek-V4-Flash
flexFP80.8× standard
$0.072$0.144$0.014$0.09
Provider API
Verified 6h ago · Changed 8h ago
More prices
  • cache write (1h TTL): $0.144
  • cache write (5m TTL): $0.09
DeepInfra
deepseek-ai/DeepSeek-V4-Flash
standardFP8
$0.09$0.18$0.018$0.113
Provider API
Verified 6h ago · Changed 8h ago
More prices
  • cache write (1h TTL): $0.18
  • cache write (5m TTL): $0.113
DeepInfra
deepseek-ai/DeepSeek-V4-Flash
priorityFP81.5× standard
$0.135$0.27$0.027$0.169
Provider API
Verified 6h ago · Changed 8h ago
More prices
  • cache write (1h TTL): $0.27
  • cache write (5m TTL): $0.169

Same weights don't always mean identical outputs: quantization, context limits and serving stacks can change quality, especially for tool calling. Prices are USD per 1M tokens; hyperscaler prices are for the reference region shown.

Monthly cost estimate

DeepInfra fp8$12.6/mo

Cache hits use each provider's cached-input price where it has one. Cache write costs aren't included.