DeepSeek V4 Flash 0731

DeepSeekKey deepseek-v4-flash-0731Also reported as deepseek/deepseek-v4-flash-0731, deepseek-v4-flash:0731, combo/deepseek-v4-flash-0731, deepseek-ai/deepseek-v4-flash-0731, hyper/deepseek-v4-flash-0731, @cf/deepseek-ai/deepseek-v4-flash-0731, /data/rongtao/models/deepseek-v4-flash-0731, nvidia/deepseek-ai/deepseek-v4-flash-0731, alicn/deepseek-v4-flash-0731, qwen-cloud/deepseek-v4-flash-0731, me/deepseek-v4-flash-0731, agent/openrouter-c/deepseek/deepseek-v4-flash-0731, ali/deepseek-v4-flash-0731, alibaba-token-plan-intl/deepseek-v4-flash-0731, alibaba-token/deepseek-v4-flash-0731, free/deepseek-v4-flash-0731, openrouter/deepseek/deepseek-v4-flash-0731, deepseek/alibaba:deepseek/deepseek-v4-flash-0731, fireworks/deepseek-v4-flash-0731, cf/@cf/deepseek-ai/deepseek-v4-flash-0731

Rank
#75
by active users
Active users
17
Down 18 from 35 active users in the previous 30 days. vs previous 30 days
Adoption
1.8%
of 967 accounts
Tokens
3.5B
cache included
Est. cost
$34
at public list prices
First seen
Jul 12, 2026
in any submission

Oct 15 – Nov 13, 2026 (UTC)967 accounts100% of tokens from an identified modelData through Nov 13, 2026, 07:12 UTCHow we count

Daily active users

Distinct accounts that used this model on each UTC day.

0 active users on Nov 13, 2026

Daily active users for DeepSeek V4 Flash 0731, Oct 15, 2026 to Nov 13, 2026. Peak 6. Use the left and right arrow keys to read each day.630
Show daily values
Daily active users and tokens for DeepSeek V4 Flash 0731
Date (UTC)Active usersTokens
Oct 15, 2026262.9M
Oct 16, 20266636.3M
Oct 17, 20265510.9M
Oct 18, 20262479.7M
Oct 19, 2026288.9M
Oct 20, 2026143.5M
Oct 21, 2026235M
Oct 22, 20263163.8M
Oct 23, 20262120.6M
Oct 24, 2026227.7M
Oct 25, 202635.4M
Oct 26, 2026279.8M
Oct 27, 2026247.2M
Oct 28, 20264103.8M
Oct 29, 20262112.6M
Oct 30, 2026469.3M
Oct 31, 202626.8M
Nov 1, 202612.9M
Nov 2, 20264122.9M
Nov 3, 2026325.8M
Nov 4, 202648.3M
Nov 5, 2026416.8M
Nov 6, 20262191.2M
Nov 7, 20263355M
Nov 8, 20264123.3M
Nov 9, 2026236.2M
Nov 10, 202611.3M
Nov 11, 202611.5M
Nov 12, 202621.8M
Nov 13, 202600

Token mix

How the 3.5B tokens split. Cache reads are usually most of an agent's tokens and are priced far below input.

  • Input6.8%
  • Output0.8%
  • Cache read92%
  • Cache write<0.1%
  • Reasoning0.2%

Agents that carry DeepSeek V4 Flash 0731

Each agent's share of this model's tokens, and how many accounts used the model through it.

  1. Hermes Agent56% of this model's tokens4active users
  2. Droid38% of this model's tokens1active users
  3. OpenCode2.0% of this model's tokens4active users
  4. DeepSeek Harness1.9% of this model's tokens1active users
  5. Claude Code1.9% of this model's tokens1active users
  6. Reasonix0.7% of this model's tokens1active users
  7. Oh My Pi0.2% of this model's tokens2active users
  8. Gajae Code<0.1% of this model's tokens1active users
  9. Senpi (OmO Native)<0.1% of this model's tokens1active users
  10. Amp<0.1% of this model's tokens1active users
  11. Cherry Studio<0.1% of this model's tokens1active users
  12. cc-mirror/curl<0.1% of this model's tokens1active users

Benchmarks

The best result for DeepSeek V4 Flash 0731 in each benchmark Tokscale tracks. Each benchmark has its own unit, run conditions and source, so compare a score only with other models on the same benchmark. Rank is among the models Tokscale matched in that benchmark.

All benchmarks
  • Epoch Capabilities IndexOverall · Epoch AI
    154.4
    Rank #51 of 158 matched modelsNo conditions reportedModel released Jul 31, 2026Source: epoch.ai (opens in a new tab)
  • GPQA DiamondReasoning · Epoch AI
    91.0%
    Rank #34 of 160 matched modelsMaxRun date Aug 2, 2026Source: epoch.ai (opens in a new tab)

    Near its ceiling: the top score is 96.5%, and Epoch AI estimates the reachable maximum at 97%. Differences between top models are small.

  • FrontierMath Tiers 1–3 (v2)Reasoning · Epoch AI
    57.5%
    Rank #47 of 90 matched modelsMaxRun date Aug 2, 2026Source: epoch.ai (opens in a new tab)

    Near its ceiling: the top score is 95.1%, and Epoch AI estimates the reachable maximum at 95%. Differences between top models are small.

  • SimpleQA VerifiedKnowledge · Epoch AI
    33.6%
    Rank #63 of 86 matched modelsMaxRun date Aug 27, 2026Source: epoch.ai (opens in a new tab)

Not evaluated: Humanity's Last Exam, Terminal-Bench 4.1, CursorBench, DeepSWE, FrontierSWE, Agent Arena, WebDev Arena. A missing result is not a score of zero.

Pricing

List prices from models.dev in US dollars per 1M tokens, retrieved Nov 12, 2026. OpenRouter is the reference price for the estimates below. It is the first provider listed; models.dev lists no price from the lab's own API.

Price at real agent usageEstimate at list prices

What 1M tokens cost at the reference list price when they split into input, output and cache the way Tokscale users' agents used them. It is not what anyone paid.

At this model's mix
$0.04
Est. per 1M tokens
92% cache reads
At the community mix
$0.04
Est. per 1M tokens
96% cache reads
3:1 input/output blend
$0.11
Per 1M tokens at list prices
No cache

Cache reads were 92% of DeepSeek V4 Flash 0731 tokens in the last 30 days and 96% of all tokens. The 3:1 blend, (3 × input + output) ÷ 4, is the usual comparison and ignores cache.

Assumptions

  • OpenRouter list prices per 1M tokens: input $0.04, output $0.30, cache read $0.04.
  • This model's mix is its own input, output, cache and reasoning tokens in the last 30 days.
  • The community mix is the tokens of every identified model in the last 30 days.
  • This is an estimate at list prices, not what anyone paid.
  • Reasoning tokens are priced as output.
  • Long-context tiers, batch pricing and discounts are not applied.
  • Token mixes come from public Tokscale submissions. Hidden accounts are excluded.
  • No cache-write price is listed, so cache writes are priced as input.

List prices by provider

  • OpenRouterReferencedeepseek/deepseek-v4-flash-0731
    $0.04 input per 1M tokens$0.30 output per 1M tokens$0.04 cache read per 1M tokensCache write not listed1.05M tokens of context
  • Vercel AI Gatewaydeepseek/deepseek-v4-flash-0731
    $0.13 input per 1M tokens$0.26 output per 1M tokens$0.03 cache read per 1M tokensCache write not listed1M tokens of context
  • Together AIdeepseek-ai/DeepSeek-V4-Flash-0731
    $0.14 input per 1M tokens$0.28 output per 1M tokens$0.03 cache read per 1M tokensCache write not listed1.05M tokens of context
  • Deep Infradeepseek-ai/DeepSeek-V4-Flash-0731
    $0.06 input per 1M tokens$0.18 output per 1M tokens$0.01 cache read per 1M tokensCache write not listed1.05M tokens of context
  • Basetendeepseek-ai/DeepSeek-V4-Flash-0731
    $0.13 input per 1M tokens$0.26 output per 1M tokens$0.03 cache read per 1M tokensCache write not listed1.05M tokens of context

Specs

From models.dev (MIT), retrieved Nov 12, 2026, as the lab or the first listed provider reports them. A dash means models.dev does not list the value.

Context window
1.05M tokens
Max output
944K tokens
Input
Text
Output
Text
Knowledge cutoff
May 2025
Release date
Jul 31, 2026
Open weights
Yes
Reasoning
Yes
Tool calling
Yes

Top public profiles

Public profiles that used DeepSeek V4 Flash 0731 most in this period, by tokens. Hidden accounts are never listed.

  1. 1Shun Kakinoki@shunkakinoki1.3Btokens—est. cost4active daysno price · 4 active days
  2. 2Benjamin Hodgens@bhodgens1Btokens—est. cost12active daysno price · 12 active days
  3. 3Alex Ozerov@BOTKooper931.6Mtokens$28est. cost29active days$28 est. cost · 29 active days
  4. 4Anthony-Hoo@Anthony-Hoo66.8Mtokens$2est. cost7active days$2 est. cost · 7 active days
  5. 5Matt Mao@ugandasniper66.4Mtokens—est. cost1active daysno price · 1 active days
  6. 6Mihael Nikolov@TheDjRider37.2Mtokens$2est. cost1active days$2 est. cost · 1 active days
  7. 7皓@hpware24.9Mtokens$2est. cost1active days$2 est. cost · 1 active days
  8. 8Baptiste@BaptisteFaisy24.2Mtokens<$1est. cost2active days<$1 est. cost · 2 active days
  9. 9SangBum Kim@Puri126.7Mtokens—est. cost2active daysno price · 2 active days
  10. 10Komal Vardhan Lolugu@KomalSrinivasan5.5Mtokens<$1est. cost1active days<$1 est. cost · 1 active days

More from DeepSeek

Other DeepSeek models in the same period, by active users.

  1. DeepSeek V4.1 Flash24% adoption230active users
  2. DeepSeek V4.1 Flash Free16% adoption150active users
  3. DeepSeek Flash12% adoption120active users
  4. DeepSeek V4.1 Flash 10178.2% adoption79active users
  5. DeepSeek V4 Pro5.5% adoption53active users
  6. DeepSeek V4 Flash5.0% adoption48active users
All models

Sources and attribution

Each score keeps its own source, unit and run conditions. Tokscale does not combine benchmarks into one score, and a missing score means the model was not evaluated, not that it scored zero.

  • Epoch AIEpoch Capabilities Index, Epoch's own runs and the external results it collects.CC BY 4.0 · Retrieved Nov 12, 2026 · Data through Nov 9, 2026Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from https://epoch.ai/benchmarks.
  • models.devSpecs and list prices per provider.MIT · Retrieved Nov 12, 2026models.dev, an open catalog of AI models (github.com/anomalyco/models.dev).
  • TokscaleActive users, tokens and token mix from public Tokscale submissions. Hidden accounts are excluded.How we count