Skip to content
Yardstick
  1. Models
  2. DeepSeek
  3. DeepSeek-V4-Pro

DeepSeek-V4-Pro

DeepSeek · Flagship · Aug 13, 2026NewOpen weights

The large text-only model of DeepSeek's V4 generation, released with open weights. DeepSeek says its newer V4.1-Flash now outperforms it.

Announcement (opens in a new tab)
Artificial Analysis Intelligence Index
36#20 of 29 · max effort
API price per 1M tokens
$1.32 / $3.96input / output · $1.98 blended
Context window
1Mup to 384K output tokens
Output speed
100 t/s22s to first answer at max effort

Benchmark results

Each row shows this model (blue) against every other model in the catalogue (grey), worst on the left, best on the right.

Overall capability

  • Artificial Analysis Intelligence Index36#20 of 29

    Max effort

    11.357.6
  • Epoch Capabilities Index155.3#12 of 16
    146.9167.3
  • LiveBench77.4%#10 of 23
    63.9%83.4%

Coding

  • Arena WebDev1,582#18 of 27

    High effort

    1,2631,814

Reasoning

  • FrontierMath (Tiers 1–3)64.6%#12 of 14

    Max effort

    53%93.7%

Knowledge & science

  • Humanity's Last Exam (Diamond)13.4%#9 of 9
    13.4%60.6%
  • GPQA Diamond91.7%#9 of 14

    Max effort

    90.2%95.8%

Human preference

  • Arena Text1,464#16 of 24

    High effort

    1,4251,525

Reasoning effort trade-offs

The same model at different thinking budgets. More effort buys a higher intelligence score at a higher cost per task and a longer wait. Measured by Artificial Analysis.

EffortScoreCost / taskFirst answerSpeed
No reasoning
20.4
$0.461.6s107 t/s
Max (headline score)
36
$0.6722s100 t/s

Specs & pricing

Developer
DeepSeek
Released
Aug 13, 2026
Availability
Generally available
API model ID
deepseek-v4-pro
Weights
Open · MIT
Parameters
1.6T (MoE)
Reasoning
Adjustable effort
Accepts
Text
Produces
Text
Context window
1,000,000 tokens
Max output
384,000 tokens
Input price
$1.32 / 1M tokens
Output price
$3.96 / 1M tokens
Cached input
$0.044 / 1M tokens

Peak-hour rates. DeepSeek charges half these prices off-peak, which covers most hours of the week.

Closest alternatives

Models from other labs nearest to DeepSeek-V4-Pro on Artificial Analysis Intelligence Index.

  • GPT-6 Luna

    OpenAI

    Intelligence 38.1 · $0.10 / $0.50

    Compare side by side

  • Qwen3.8-27B

    Alibaba

    Intelligence 33.7 · $0.50 / $3

    Compare side by side

  • Mistral Large 4

    Mistral AI

    Intelligence 38.4 · $1.36 / $4.18

    Compare side by side

Sources

  • Announcement (opens in a new tab)
  • Model documentation (opens in a new tab)
  • Pricing (opens in a new tab)
Yardstick

A plain reading of how today’s SI models compare. Scores are taken from each benchmark’s own leaderboard wherever one exists; anything a lab reported about its own model is marked.

Data as of Oct 8, 2026.

Explore

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Methodology & sources

Sources

  • Artificial Analysis (opens in a new tab)
  • Arena (opens in a new tab)
  • Epoch AI Benchmarking Hub (opens in a new tab)
  • ARC Prize (opens in a new tab)
  • Terminal-Bench (opens in a new tab)
  • Scale AI Labs (opens in a new tab)

Yardstick is an independent reference and is not affiliated with, sponsored by or endorsed by any lab listed. Product names and logos are trademarks of their respective owners and are used only to identify them. Benchmarks are imperfect proxies: check a model on your own task before you commit to it.

  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs
  • Models
  • Compare
  • Benchmarks
  • Value
  • Labs