ModelPriceWatch.com
Last scan 2026-08-21 Models tracked 220 Providers 31 Cheapest paid Granite 4.0 H Micro $0.017/Mtok in Every price links to its source

DeepSeek V4 Flash Vision Exp

by DeepSeek

Preview new · 0d budget cheap tier
Availability: DeepSeek labels this an experimental model in its own API docs and the SKU name carries the -exp suffix, so it is tracked as Preview rather than Current. It is nonetheless publicly callable and priced on DeepSeek's standard rate card, listed as a full column beside deepseek-v4-flash and deepseek-v4-pro with its own 2500-request concurrency limit.

Today's price · per 1M tokens

Input

$0.440

per 1M tokens

Output

$1.32

per 1M tokens

Blended

$0.660

blended $/1M — 3:1 weighted input:output

Cached input

$0.014

3% of input — prompt caching

How this price is scoped: DeepSeek bills this model at TWO rates depending on the hour, on the same schedule and at the same numbers as deepseek-v4-flash. The tracked figure is the PEAK rate: $0.44 input / $0.014 cached input / $1.32 output per 1M tokens. Off-peak is exactly half: $0.22 / $0.007 / $0.66. Peak hours are 01:00-04:00 and 06:00-10:00 UTC; every other hour is off-peak. Peak is the headline because DeepSeek defines off-peak as half of peak rather than the other way round, so peak is the list rate. Image input bills at the same per-token rates once the image is tokenized; DeepSeek's rate card prints no separate per-image charge.

Source: official DeepSeek pricing · read Aug 21, 2026 MODELPRICEWATCH.COM · 2026-08-21

Price receipt

We read DeepSeek's own pricing page on Aug 21, 2026 and found deepseek-v4-flash-vision-exp listed with a price on the same column — that read is where the number above comes from, recorded as a new model. We keep the page text we read; its content hash is c137a331b2.

Overview

Vision-enabled variant of DeepSeek V4 Flash, adding image input while matching the base model on text. Same 1M-token context, 384K max output and prompt caching as V4 Flash, and priced identically: $0.44/$1.32 per 1M tokens at peak (01:00-04:00 and 06:00-10:00 UTC); half that in the 17 off-peak hours.

Capabilities

struck through = not supported
Input 2/5
Text ✓ Image ✓ Audio Video PDF
Output 1/5
Text ✓ Image Audio Video Embedding
Features 4/9
Prompt caching ✓ Reasoning Coding Fast inference ✓ Long context ✓ Open weights Multimodal ✓ Web search Realtime

Specifications

Provider
DeepSeek
Context window
1M tokens
Modality
text, image
Parameters
Proprietary
Open source
No — proprietary
Released
Aug 21, 2026
Status
Preview
Last updated
Aug 21, 2026
Tags
fastmultimodalcachingbudget

Availability verified: Aug 21, 2026 — listed on DeepSeek's own page