LLM Pricing
ModelsLeaderboardsToolsProvidersReleasesArticles

© 2026 LLM Pricing

About
·Contact
·Privacy
·RSS
← Model list

DeepSeek V4 Flash 0731

deepseek·deepseek/deepseek-v4-flash-0731·BETA·Open weights·deepseek-flash series
Official DeepSeek V4 Flash release with enhanced agentic capabilities and integrated DSpark speculative decoding

Specs & pricing

Input / output per 1M tokens
Official price
—
Lowest paid·JoinGonkaGateway
$0.0062 / $0.019
Blended $0.05 · 14.8× spread
2 more $0 channels
Context
524,288
Output limit
384,000
Knowledge cutoff
2025-05
Released / updated
2026-07-31 / 2026-07-31
Capabilities
✓ Reasoning✓ Tool use✓ Structured output✓ TemperatureAttachments
⚠ Providers report capability flags inconsistently
Modalities
Text
License
MIT
Weights
Hugging Face

DeepSeek V4 Flash 0731 is offered by 92 providers on this page. Public prices are shown for 86 of them. No official API price is listed for it; the prices below come from third-party channels. The lowest paid channel is SAGG at $0.04 / $0.079 per 1M. That channel is a third-party gateway, so confirm its availability and rate limits before depending on it. Prices vary about 15× across the providers that host it. It also has 2 free ($0) channels and 4 covered by a paid subscription; free tiers usually carry rate limits, and subscription-covered access bills $0 per token only after the subscription fee.

Median output speed is 125 tokens per second, with 1.41s to the first token.

The context window is 524,288 tokens at the reference host, but hosts report different limits, from 200,000 to 1,050,000, so the usable window depends on the provider you pick. It supports reasoning, tool use, and structured output. The weights are open, so it can also be self-hosted or served through a gateway of your choice. Providers report the capability flags inconsistently, so verify a specific feature against the host you plan to use. Its training knowledge cuts off at 2025-05.

Quality & performanceArtificial Analysis · Intelligence Index v5 · rep. tier Max

Intelligence—
Value—
Output speed125 tok/s
TTFT1.41 s

Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.

Available at 92 providers86 with public prices · 2 free · 4 subscription-covered

  • ProviderInput / Output
  • Umans AI Coding PlanGateway
    umans-deepseek-v4-flash-0731
    Subscription
    Cache read $0Context 1,048,576 ⚠Output limit 393,215deprecated
  • AMDCloud
    DeepSeek-V4-Flash
    Free
    Cache read $0Context 1,048,576 ⚠Output limit 384,000
  • SCNet Token PlanGateway
    DeepSeek-V4-Flash-0731
    Subscription
    Cache read $0Context 1,000,000 ⚠Output limit 384,000
  • Alibaba Token PlanGateway
    Subscription
    Cache read $0Context 1,000,000 ⚠Output limit 384,000
  • Alibaba Token Plan (China)Gateway
    Subscription
    Cache read $0Context 1,000,000 ⚠Output limit 384,000
  • OpenCodeGateway
    deepseek-v4-flash-free
    Free
    Cache read $0Context 200,000 ⚠Output limit 128,000deprecated
  • JoinGonkaGateway
    deepseek-ai/DeepSeek-V4-Flash-0731
    $0.0062 ⚠ / $0.019
    20×+ off the median, excluded from lowest price
    Context 380,000 ⚠Output limit 32,768
  • SAGGGateway
    deepseek-ai/DeepSeek-V4-Flash-0731
    $0.04 / $0.079
    Context 400,000 ⚠Output limit 16,000
  • engyGateway
    $0.045 / $0.09
    Cache read $0.009Context 1,048,576 ⚠Output limit 128,000
  • AdaptionGateway
    deepseek-v4-flash
    $0.044 / $0.13
    Cache read $0.0014Context 1,048,576 ⚠Output limit 384,000
ProviderTierInputOutputCache readCache writeContextOutput limitStatus
Umans AI Coding Plan
umans-deepseek-v4-flash-0731
GatewaySubscription$0$01,048,576 ⚠393,215deprecated
AMD
DeepSeek-V4-Flash
CloudFree$0—1,048,576 ⚠384,000
SCNet Token Plan
DeepSeek-V4-Flash-0731
GatewaySubscription$0—1,000,000 ⚠384,000
Alibaba Token PlanGatewaySubscription$0$01,000,000 ⚠384,000
Alibaba Token Plan (China)GatewaySubscription$0$01,000,000 ⚠384,000
OpenCode
deepseek-v4-flash-free
GatewayFree$0—200,000 ⚠128,000deprecated
JoinGonka
deepseek-ai/DeepSeek-V4-Flash-0731
Gateway$0.0062 ⚠
20×+ off the median, excluded from lowest price
$0.019——380,000 ⚠32,768
SAGG
deepseek-ai/DeepSeek-V4-Flash-0731
Gateway$0.04$0.079——400,000 ⚠16,000
engyGateway$0.045$0.09$0.009—1,048,576 ⚠128,000
Adaption
deepseek-v4-flash
Gateway$0.044$0.13$0.0014—1,048,576 ⚠384,000

Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.

Reasoning control

Toggle (on / off)effort = loweffort = higheffort = maxeffort = noneeffort = minimaleffort = mediumeffort = xhighbudget_tokens ≥ 1 ≤ 393,216

Interleaved thinking (reasoning between tool calls) is declared by 42 of 92 providers.

2 / 92 providers expose no reasoning control (reasoning_options: []).

Your usage cost

Common scenarios
1JoinGonka$2.16
2engy$9.18
3Adaption$10.39
4SAGG$11.91
5NanoGPT$13.56
6NanoGPT$13.56

6 more channels offer $0 (Umans AI Coding Plan, AMD, SCNet Token Plan etc.); free tiers usually have rate limits and no SLA, excluded from ranking.

The cheapest paid channel is the only channel.

Price historyone sample accumulated per data sync

Input listOutput listMin blended
$0.11$02026-08-052026-11-15
2026-11-15Input list —Output list —Min blended $0.05

Benchmark9 items

NameConditionsScoreMetricSource
Terminal-Benchvariant: max · v2.182.7pass@1Source ↗
NL2Repoharness: DeepSeek Harness minimal mode · variant: max effort54.2resolve rateSource ↗
CyberGymharness: DeepSeek Harness minimal mode · variant: max effort76.7scoreSource ↗
DeepSWEharness: DeepSeek Harness minimal mode · variant: max effort54.4resolve rateSource ↗
Toolathlon-Verifiedharness: DeepSeek Harness minimal mode · variant: max effort70.3scoreSource ↗
Agents' Last Examharness: DeepSeek Harness minimal mode · variant: max effort25.2scoreSource ↗
AutomationBenchvariant: max effort · dataset: public25.1success rateSource ↗
DSBench-FullStackvariant: max effort · dataset: internal68.7scoreSource ↗
DSBench-Hardvariant: max effort · dataset: internal59.6scoreSource ↗

The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.

Related models

DeepSeek V4.1 Flash 1017same series$0.15 / $0.60DeepSeek: DeepSeek Flash Latestsame series—DeepSeek V4.1 Flashsame series—DeepSeek V4 Flash 0423same lab—DeepSeek V4 Flash Latestsame lab—
Data partly from models.dev (MIT)