Compare / DeepSeek-V4-Flash (0731) vs DeepSeek-V4-Pro (0423)

DeepSeek-V4-Flash (0731) vs DeepSeek-V4-Pro (0423)

Which is cheaper, which is more capable, and which does better at each kind of work, from 10 results on 2 sources that measured both models.

DeepSeekDeepSeek-V4-Flash (0731) Released 31 Jul 2026 · $0.13 in · $0.40 out
DeepSeekDeepSeek-V4-Pro (0423) Released 24 Apr 2026 · $1.42 in · $2.83 out
DeepSeek-V4-Flash (0731) wins 7DeepSeek-V4-Pro (0423) wins 3

Of the 10 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 16 November 2026 (catalogue release 2026-11-16-49bdfa045dee).

DeepSeek-V4-Flash (0731) or DeepSeek-V4-Pro (0423)? The short answer

Pick DeepSeek-V4-Flash (0731): it scores higher (Intelligence Index 34.3 against 30.4) and costs less.

  • DeepSeek-V4-Flash (0731) is cheaper: $0.20 against $1.77 per million tokens, 8.9 times less.
  • DeepSeek-V4-Flash (0731) scores higher on the Artificial Analysis Intelligence Index: 34.3 against 30.4.
  • DeepSeek-V4-Flash (0731) is newer: released 31 Jul 2026, against 24 Apr 2026 for DeepSeek-V4-Pro (0423).
  • On third-party benchmarks, on the results both models have: DeepSeek-V4-Flash (0731) does better for agents and tool use and reasoning and knowledge; DeepSeek-V4-Pro (0423) for sticking to the facts.

Head to head

34.3
Intelligence Index
30.4
$0.20
Price per million tokens, blendedlower is better
$1.77
1,048,576
Context window, tokens
1,048,576

Every fact side by side

DeepSeek-V4-Flash (0731)DeepSeek-V4-Pro (0423)
Price per million tokens$0.13 in · $0.40 out$1.42 in · $2.83 out
Blended (3 in : 1 out)$0.20$1.77
Cheapest host, blended$0.088 (Relace, FP4)$1.05 (Relace, FP4)
Intelligence Index34.330.4
Spring Prompt overall–28
Output speed, tokens a second––
Context window1,048,576 tokens1,048,576 tokens
Released31 Jul 2026 (newer)24 Apr 2026
WeightsOpenOpen
Developer based inChinaChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Decks from an analysisTurning a finished analysis into a deck you could present as it is not measured 32 of 32
Marketing planningPlanning a year of ad spend without overspending not measured 26 of 32
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 43 of 159 92 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 47 of 187 51 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality not measured 30 of 169
Sticking to the factsSummarising without inventing things, and factual answers 105 of 139 56 of 139
Speed under pressureGood decisions against a real clock (fast chess) not measured 36 of 36

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureDeepSeek-V4-Flash (0731)DeepSeek-V4-Pro (0423)
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 44 · 53 of 249 measured 34.3max reasoning 30.4max reasoning
GPQA Diamond% of questions, higher is better 35 · 37 of 232 measured 90.8%max reasoning 90.5%high reasoning
Humanity's Last Exam% of questions, higher is better 52 · 57 of 250 measured 38.6%max reasoning 37.5%max reasoning
SimpleQA Verified (Epoch AI) reported by SimpleQA Verified (Epoch AI) · 1 result
Correct answers% of questions, higher is better 54 · 32 of 76 measured 33.6%max reasoning 47.0%max reasoning
Show all 10 results (6 more)
MeasureDeepSeek-V4-Flash (0731)DeepSeek-V4-Pro (0423)
Artificial Analysis reported by Artificial Analysis · 6 results
Terminal-Bench 2.1% of tasks, higher is better 26 · 42 of 131 measured 78.7%max reasoning 64.8%high reasoning
Terminal-Bench 4.0% of tasks, higher is better 40 · 32 of 120 measured 12.1%max reasoning 14.6%max reasoning
τ-bench banking% of tasks, higher is better 21 · 35 of 121 measured 39.4%max reasoning 30.1%max reasoning
Artificial Analysis Coding Indexindex score, higher is better 29 · 40 of 130 measured 69.1max reasoning 59.4max reasoning
Long-context reasoning (AA-LCR)% of questions, higher is better 47 · 86 of 241 measured 79.7%max reasoning 74.7%max reasoning
SciCode% of problems, higher is better 53 · 50 of 121 measured 50.3%max reasoning 50.8%max reasoning

Quick answers

Is DeepSeek-V4-Flash (0731) better than DeepSeek-V4-Pro (0423)?

Pick DeepSeek-V4-Flash (0731): it scores higher (Intelligence Index 34.3 against 30.4) and costs less.

Which is cheaper, DeepSeek-V4-Flash (0731) or DeepSeek-V4-Pro (0423)?

DeepSeek-V4-Flash (0731) is cheaper: $0.20 against $1.77 per million tokens, 8.9 times less.

Which is better for agents and tool use, DeepSeek-V4-Flash (0731) or DeepSeek-V4-Pro (0423)?

DeepSeek-V4-Flash (0731) ranks 43 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 92 for DeepSeek-V4-Pro (0423).

Which is better for reasoning and knowledge, DeepSeek-V4-Flash (0731) or DeepSeek-V4-Pro (0423)?

DeepSeek-V4-Flash (0731) ranks 47 of 187 models on sale for hard questions across science, maths and general knowledge, against 51 for DeepSeek-V4-Pro (0423).

Which is better for sticking to the facts, DeepSeek-V4-Flash (0731) or DeepSeek-V4-Pro (0423)?

DeepSeek-V4-Pro (0423) ranks 56 of 139 models on sale for summarising without inventing things, and factual answers, against 105 for DeepSeek-V4-Flash (0731).

Run DeepSeek-V4-Flash (0731) and DeepSeek-V4-Pro (0423) on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.