Skip to content

DeepSeek V4.1 Flash

text + image → textHugging Face OpenRouter Cost calculator →

Best price · blended / 1M

$0.0887

at Baidu

Input / output
$0.0507 / $0.203
Providers
31
Spread
27×
7 days
−23.8%
30 days
+73.2%
Context
1M
Prices per 1M tokens, blended: Baidu $0.0887, Alibaba $0.158, DeepInfra (fp8/flex) $0.168, Wafer (us) $0.18, DeepInfra $0.21, Novita $0.21, River $0.21, Relace $0.215, InferenceNet $0.227, InferenceNet (us) $0.227, AtlasCloud $0.236, Reka $0.25, StreamLake $0.257, CoreWeave $0.26, OpenInference $0.27, Logit $0.275, Cloudflare $0.30, Wafer $0.305, Cohere $0.311, DeepInfra (fp8/priority) $0.315, GMICloud $0.315, Ionstream $0.325, Mancer 2 $0.347, Fireworks (spot) $0.368, DigitalOcean $0.394, Makora $0.488, Crusoe $0.518, BaseTen $0.525, BaseTen (us) $0.525, Fireworks $0.525, Google (global) $0.525, Modal $0.525, Nebius $0.525, Parasail $0.525, SiliconFlow $0.525, Together $0.525, Venice $0.525, Fireworks (us) $0.788, Parasail (fast) $0.788, Venice (ultrafast) $5.25.

Providers

  • BaiduCheapest
    $0.0887
    $0.0507 in · $0.203 out
    −55% vs ref
    1M ctx100% uptimefp8·−83% discount
  • $0.158
    $0.09 in · $0.36 out
    −19% vs ref
    1M ctx100% uptime
  • fp8/flex
    $0.168
    $0.112 in · $0.336 out
    −14% vs ref
    1M ctx95.7% uptimefp8·−30% discount
  • $0.18
    $0.051 in · $0.567 out
    −7.7% vs ref
    1M ctx100% uptimemxfp4
  • $0.21
    $0.14 in · $0.42 out
    +7.7% vs ref
    1M ctx100% uptimefp8·−30% discount
  • $0.21
    $0.12 in · $0.48 out
    +7.7% vs ref
    1M ctx100% uptimefp8·−60% discount
  • $0.21
    $0.12 in · $0.48 out
    +7.7% vs ref
    1M ctx99.9% uptime
  • $0.215
    $0.12 in · $0.50 out
    +10% vs ref
    1M ctx100% uptimefp4
  • $0.227
    $0.103 in · $0.60 out
    +17% vs ref
    1M ctx100% uptimefp8
  • $0.227
    $0.103 in · $0.60 out
    +17% vs ref
    1M ctx100% uptimefp8
  • $0.236
    $0.135 in · $0.54 out
    +21% vs ref
    1M ctx100% uptimefp8
  • $0.25
    $0.12 in · $0.64 out
    +28% vs ref
    1M ctx100% uptime−20% discount
  • $0.257
    $0.147 in · $0.588 out
    +32% vs ref
    1M ctx100% uptimefp8·−51% discount
  • $0.26
    $0.13 in · $0.65 out
    +33% vs ref
    1M ctx100% uptimefp8
  • $0.27
    $0.06 in · $0.90 out
    +38% vs ref
    1M ctx100% uptimefp4
  • $0.275
    $0.10 in · $0.80 out
    +41% vs ref
    1M ctxfp8
  • $0.30
    $0.10 in · $0.90 out
    +54% vs ref
    1M ctx99.6% uptime
  • $0.305
    $0.19 in · $0.65 out
    +56% vs ref
    1M ctx100% uptimemxfp4
  • $0.311
    $0.195 in · $0.66 out
    +60% vs ref
    1M ctx100% uptimefp8
  • fp8/priority
    $0.315
    $0.21 in · $0.63 out
    +62% vs ref
    1M ctx99.9% uptimefp8·−30% discount
  • $0.315
    $0.18 in · $0.72 out
    +62% vs ref
    1M ctx100% uptimefp8·−40% discount
  • $0.325
    $0.20 in · $0.70 out
    +67% vs ref
    1M ctx100% uptime
  • $0.347
    $0.213 in · $0.75 out
    +78% vs ref
    1M ctx100% uptimefp8
  • $0.368
    $0.21 in · $0.84 out
    +88% vs ref
    1M ctx98.8% uptime
  • $0.394
    $0.225 in · $0.90 out
    +102% vs ref
    1M ctx100% uptime
  • $0.488
    $0.26 in · $1.17 out
    +150% vs ref
    1M ctx90.5% uptimefp8
  • $0.518
    $0.29 in · $1.20 out
    +165% vs ref
    1M ctx98.0% uptimefp8
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • global
    $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptimefp8
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptimefp8
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptimefp8
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptime
  • $0.525
    $0.30 in · $1.20 out
    +169% vs ref
    1M ctx100% uptimefp8
  • $0.788
    $0.45 in · $1.80 out
    +304% vs ref
    1M ctx100% uptime
  • $0.788
    $0.45 in · $1.80 out
    +304% vs ref
    1M ctx100% uptimefp8
  • ultrafast
    $5.25
    $3.00 in · $12 out
    +2592% vs ref
    1M ctx100% uptimefp8
Blended = (3 × input + output) / 4. Ref $0.195 = median of cheapest 20%.

Price history

USD per 1M tokens · each step is a price change

Prices in USD / 1M tokens.

Long-context pricing

Offered byPrompt sizeInputOutput
Alibababase$0.09$0.36
> 0$0.18$0.72
> 0$0.09$0.36

Other pricing

Offered byCache read
Baidu$0.00101
Alibaba$0.0018
DeepInfra (fp8/flex)$0.00336
Wafer (us)$0.018
DeepInfra, Fireworks (spot)$0.0042
Novita, River$0.0024
Relace, Cohere$0.003
InferenceNet, InferenceNet (us) +1$0.008
AtlasCloud$0.0135
StreamLake$0.00294
CoreWeave, OpenInference$0.002
Logit$0.10
Cloudflare, Google (global) +1$0.03
Wafer$0.012
DeepInfra (fp8/priority)$0.0063
GMICloud$0.0036
Ionstream$0.02
Mancer 2$0.05
DigitalOcean, Fireworks +3$0.006
Makora$0.0058
Crusoe, BaseTen +1$0.007
Nebius$0.01
Venice$0.0075
Fireworks (us), Parasail (fast)$0.009
Venice (ultrafast)$0.06

Changes

  • Cache read price,
    $0.0036→$0.0018−50.0%
  • Output price,
    $0.72→$0.36−50.0%
  • Input price,
    $0.18→$0.09−50.0%
  • Discount price,
    —→83%—
  • Cache read price,
    $0.006→$0.00101−83.1%
  • Output price,
    $1.20→$0.203−83.1%
  • Input price,
    $0.30→$0.0507−83.1%
  • Cache read price,
    $0.003→$0.002−33.3%
  • Output price,
    $1.00→$0.90−10.0%
  • Cache read price,
    $0.0018→$0.0036+100%
  • Output price,
    $0.36→$0.72+100%
  • Input price,
    $0.09→$0.18+100%
  • Cache read price,
    $0.005→$0.003−40.0%
  • Output price,
    $1.20→$1.00−16.7%
  • Input price,
    $0.09→$0.06−33.3%

About

The cheapest provider for DeepSeek V4.1 Flash is Baidu at $0.0887 per 1M tokens (blended), charging $0.0507 per 1M input and $0.203 per 1M output tokens. 31 providers serve it; the most expensive charges 27× the typical price. Its best price rose 73.2% over the last 30 days. Context window: 1M tokens. Prices tracked since Oct 3, 2026.

DeepSeek V4.1 Flash is a sparse mixture-of-experts model from DeepSeek, and the first built on the company's Causal Encoder-Decoder (CED) architecture. It activates 8B parameters on input and 16B on...
Released
Sep 10, 2026
First tracked
Oct 3, 2026
Canonical slug
deepseek/deepseek-v4.1-flash-20260910
Weights
deepseek-ai/DeepSeek-V4.1-Flash