DeepSeek V4 Flash API pricing

9 providers serve DeepSeek V4 Flash. DeepSeek V4 Flash is an open-weight model (MIT) with 284B total parameters, 13B active per token (mixture-of-experts), up to 1049K context. Prices below are per 1M tokens, cheapest input first (self-hosting math further down).

Refreshed 2026-08-23 · list prices from provider APIs

Input floor $0.050/1M at OpenRouter Output floor $0.101/1M at OpenRouter Find your workload winner →

Provider	Input $/1M	Output $/1M	Cache read $/1M	Context
OpenRouter marketplace quote	$0.050	$0.101output floor	$0.010	1049K	Use
Pinstripes	$0.100	$0.200	-	164K
Fireworks AI	$0.140	$0.280	$0.028	1049K	Use
Tensormesh	$0.140	$0.280	-	33K
Tencent	$0.140	$0.280	$0.0028	1000K
Azure AI Foundry	$0.190	$0.510	-	1000K	Use
Alibaba Model Studio	$0.200	$0.400	$0.040	1000K	Use
Libertai	$0.250	$1.75	-	200K
DeepSeek	$0.440	$1.32	$0.014	1000K	Use

The cheapest way to run DeepSeek V4 Flash

On an illustrative 70M-input / 30M-output monthly workload, today's lowest listed cost is roughly $6.54/month, at OpenRouter. Input-heavy, output-heavy and cache-heavy workloads can produce different winners. Use the live mix calculator below rather than combining floors from two different providers. On output-token cost alone, the API floor remains below our self-hosting estimate of ~$0.209/1M. Details below. Run your own numbers in the breakeven calculator, or read the breakeven math.

Your actual monthly cost

Uses today's list prices

Workload preset Input tokens / month (millions) Output tokens / month (millions) Input served from cache %

Cache read uses the listed cache rate where available; otherwise it falls back to normal input price. Batch, write-cache, volume and negotiated discounts are excluded.

Or run it yourself: DeepSeek V4 Flash self-hosting economics

DeepSeek V4 Flash is open-weight (MIT), so the API price above competes with the GPU-hour market. 284B total parameters (MoE, ~13B active per token) needs roughly 391 GB VRAM at FP8 or 196 GB at INT4, KV-cache headroom included.

No single GPU fits DeepSeek V4 Flash at FP8 (391 GB needed), so it needs a multi-GPU node. Cheapest tracked option: 3× H200 NVL (141 GB each) at roughly $1.50/hr total (RunPod). Rent 3× H200 NVL at RunPod → INT4 quantization drops the requirement to 196 GB. Tune it in the calculator.

Breakeven estimate: a well-batched vLLM deployment on 3× H200 NVL ($1.50/hr) at ~3981 aggregate tok/s and 50% utilization works out to roughly $0.209 per 1M output tokens, versus $0.101 via the cheapest output-token API (OpenRouter). The API is cheaper unless your utilization or throughput beats these assumptions. Planning estimate: throughput varies with hardware, quantization, batch size and context. Tune it in the calculator.

Related models

Model	Cheapest in $/1M	Cheapest out $/1M	Providers
DeepSeek R1	$0.200	$0.400	17
DeepSeek V3	$0.200	$0.200	16
DeepSeek R1 Distill Llama 70B	$0.200	$0.375	10
DeepSeek V3.2	$0.269	$0.400	9
DeepSeek-V3.1	$0.270	$1.00	9
DeepSeek V4 Pro	$0.435	$0.870	6

FAQ

What is the cheapest API for DeepSeek V4 Flash?

As of 2026-08-23, the lowest input price for DeepSeek V4 Flash is OpenRouter at $0.050 per 1M input tokens. The lowest output price is OpenRouter at $0.101 per 1M output tokens. The cheapest provider for a real workload depends on its input/output mix. The next-lowest input price is Pinstripes at $0.100, a 99% difference.

How much VRAM do you need to self-host DeepSeek V4 Flash?

DeepSeek V4 Flash has 284B parameters, so plan for roughly 391 GB of VRAM at FP8 or 196 GB at INT4/AWQ, KV-cache headroom included. That exceeds a single GPU. A typical node is 3× H200 NVL.

← All models · GPU rental prices · Breakeven calculator · Get this data as JSON