DeepSeek V4 Flash 0731 vs DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 leads DeepSeek V4 Flash 0731 on coding, 56.1 to 51.8 on the Vibe Coding Index, and DeepSeek V4 Flash 0731 costs 7.1x less per output token ($0.28 vs $1.98 per million). Updated October 21, 2026; 45 benchmarks have a result for both.
DeepSeek V4 Flash 0731 or DeepSeek V4 Pro 0813: which to pick
Pick DeepSeek V4 Pro 0813 for coding: it scores 56.1 on the Vibe Coding Index against 51.8 for DeepSeek V4 Flash 0731.
Pick DeepSeek V4 Flash 0731 for output tokens 7.1x cheaper ($0.28 vs $1.98 per million) and a lower cost per solved task ($0.01 against $0.10).
- Shared results
- 45
- DeepSeek V4 Flash 0731 only
- 10
- DeepSeek V4 Pro 0813 only
- 33
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 scores 56.1 against 51.8.
4 and 5 benchmarks behind the scores
Intelligence
AA Intelligence Index
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 leads, 36.0 against 34.3.
Same board, both measured
Agentic work
AA Agentic Index
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 leads, 41.3 against 41.0.
Same board, both measured
Reasoning
GPQA Diamond
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 leads, 92.4% against 89.9%.
Same board, both measured
Knowledge
MMLU Pro
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 leads, 87.0% against 86.2%.
Same board, both measured
Web design
Design Arena: websites
No call
Only DeepSeek V4 Flash 0731 has a Design Arena: websites result, 1238.
Directional only
General
Vals Index
No call
Only DeepSeek V4 Pro 0813 has a Vals Index result, 53.0%.
Directional only
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 $0.0003, DeepSeek V4 Pro 0813 $0.0016 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 $0.01, DeepSeek V4 Pro 0813 $0.10.
Published attempt cost and pass rate
At a glance
- Coding index
- DeepSeek V4 Flash 0731 51.8 from 4 benchmarks, DeepSeek V4 Pro 0813 56.1 from 5 benchmarks.
- Widest category
- Coding, 5.5 points apart on average over 15 shared benchmarks.
- Widest single gap
- SimpleQA Verified: DeepSeek V4 Flash 0731 33.6% against DeepSeek V4 Pro 0813 52.9%, 19.3 points apart on a 0 to 100 scale.
- Output price
- DeepSeek V4 Flash 0731 $0.28 against DeepSeek V4 Pro 0813 $1.98 per million output tokens, 7.1x apart.
- Context window
- Both take 1.05M tokens.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | DeepSeek V4 Flash 0731 | Bars | DeepSeek V4 Pro 0813 | Ties |
|---|---|---|---|---|
| Coding | 3 | 11 | 1 tie | |
| Intelligence | 0 | 2 | ||
| Agentic work | 2 | 3 | 1 tie | |
| Reasoning | 1 | 7 | ||
| Math | 1 | 3 | ||
| Knowledge | 0 | 6 | ||
| General | 1 | 3 | ||
| All shared boards | 8 | 35 | 2 ties |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index311
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1582 | Only DeepSeek V4 Pro 0813 measured |
| AA Coding Index | 69.1 | 68.8 | DeepSeek V4 Flash 0731 |
| LiveCodeBench (Vals AI) | 87.3% | 87.5% | DeepSeek V4 Pro 0813 |
| SWE-bench Verified (Vals AI) | 88.8% | 96.4% | DeepSeek V4 Pro 0813 |
| Terminal-Bench 4 (Vals AI) | 18.7% | 14.1% | DeepSeek V4 Flash 0731 |
| Vibe Code Bench (Vals AI) | 74.7% | 82.3% | DeepSeek V4 Pro 0813 |
| Code migration (Vals AI) | 38.6% | 41.5% | DeepSeek V4 Pro 0813 |
| ProgramBench (Vals AI) | 0.0% | 0.0% | Tie |
| IOI (Vals AI) | 32.7% | 51.6% | DeepSeek V4 Pro 0813 |
| Terminal-Bench 2.1 (Vals AI) | 67.0% | 54.7% | DeepSeek V4 Flash 0731 |
| Vibe Code Bench 1-100 (Vals AI) | Not measured | 17.5% | Only DeepSeek V4 Pro 0813 measured |
| SkillsBench (Vals AI) | 50.7% | 53.8% | DeepSeek V4 Pro 0813 |
| CyberBench (Vals AI) | 72.1% | 72.9% | DeepSeek V4 Pro 0813 |
| SciCode | 50.3% | 51.0% | DeepSeek V4 Pro 0813 |
| FrontierCode | 18.8% | 28.5% | DeepSeek V4 Pro 0813 |
| WeirdML | 63.0% | 66.2% | DeepSeek V4 Pro 0813 |
| ALE-Bench | 1306.1 | 1403.2 | DeepSeek V4 Pro 0813 |
| LMArena Text: coding | Not measured | 1469 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: software and IT | Not measured | 1462 | Only DeepSeek V4 Pro 0813 measured |
| LMArena WebDev: React | Not measured | 1588 | Only DeepSeek V4 Pro 0813 measured |
| LMArena WebDev: HTML | Not measured | 1566 | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 18 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| AA Intelligence Index | 34.3 | 36.0 | DeepSeek V4 Pro 0813 |
| Epoch Capabilities Index | 154.8 | 155.6 | DeepSeek V4 Pro 0813 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index23
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| AA Agentic Index | 41.0 | 41.3 | DeepSeek V4 Pro 0813 |
| SREBench (Vals AI) | Not measured | 1.9% | Only DeepSeek V4 Pro 0813 measured |
| Finance Agent (Vals AI) | 49.5% | 50.4% | DeepSeek V4 Pro 0813 |
| Tax Agent Bench (Vals AI) | 28.1% | 26.8% | DeepSeek V4 Flash 0731 |
| Legal Research Bench (Vals AI) | 30.3% | 40.9% | DeepSeek V4 Pro 0813 |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 10.8% | Only DeepSeek V4 Pro 0813 measured |
| Excel Modeling Benchmark (Vals AI) | 57.0% | 52.8% | DeepSeek V4 Flash 0731 |
| BioMysteryBench (Vals AI) | 64.4% | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Terminal-Bench-Science (Vals AI) | 4.3% | 4.3% | Tie |
| APEX-Agents | Not measured | 47.3% | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 31 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index17
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| GPQA Diamond | 89.9% | 92.4% | DeepSeek V4 Pro 0813 |
| GPQA Diamond (Artificial Analysis run) | 90.8% | 92.8% | DeepSeek V4 Pro 0813 |
| Humanity's Last Exam (Artificial Analysis run) | 38.6% | 41.0% | DeepSeek V4 Pro 0813 |
| GPQA Diamond (Epoch AI run) | 91.0% | 91.7% | DeepSeek V4 Pro 0813 |
| ARC-AGI | 89.0% | 90.5% | DeepSeek V4 Pro 0813 |
| ARC-AGI-2 | 61.4% | 61.3% | DeepSeek V4 Flash 0731 |
| SimpleBench | 61.1% | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Chess puzzles | 32.0% | 46.0% | DeepSeek V4 Pro 0813 |
| Mystery game puzzles | 34.0% | 43.0% | DeepSeek V4 Pro 0813 |
| LMArena Text: hard prompts | Not measured | 1458 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: expert | Not measured | 1459 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: hard prompts in English | Not measured | 1463 | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 10 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 94.4% | 98.6% | DeepSeek V4 Pro 0813 |
| FrontierMath, tiers 1 to 3 | 57.5% | 64.6% | DeepSeek V4 Pro 0813 |
| FrontierMath, tier 4 | 24.4% | 26.8% | DeepSeek V4 Pro 0813 |
| ProofBench | 56.0% | 50.0% | DeepSeek V4 Flash 0731 |
| LMArena Text: math | Not measured | 1441 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: mathematical work | Not measured | 1443 | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| MMLU Pro | 86.2% | 87.0% | DeepSeek V4 Pro 0813 |
| LegalBench (Vals AI) | 77.7% | 82.4% | DeepSeek V4 Pro 0813 |
| CorpFin (Vals AI) | 61.9% | 65.4% | DeepSeek V4 Pro 0813 |
| TaxEval (Vals AI) | 70.7% | 73.1% | DeepSeek V4 Pro 0813 |
| MedCode (Vals AI) | 41.4% | 42.5% | DeepSeek V4 Pro 0813 |
| SimpleQA Verified | 33.6% | 52.9% | DeepSeek V4 Pro 0813 |
| LMArena Text: science | Not measured | 1457 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: medicine and healthcare | Not measured | 1446 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: legal and government | Not measured | 1459 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: business and finance | Not measured | 1442 | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 8 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| Design Arena: websites | 1238 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: UI components | 1235 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: game dev | 1208 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: data viz | 1180 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: 3D | 1209 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: all code categories | 1230 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: SVG | 1185 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Design Arena: ASCII art | 1090 | Not measured | Only DeepSeek V4 Flash 0731 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| Vals Index | Not measured | 53.0% | Only DeepSeek V4 Pro 0813 measured |
| MedScribe (Vals AI) | 80.4% | 80.2% | DeepSeek V4 Flash 0731 |
| Long context reasoning (AA-LCR) | 79.7% | 80.3% | DeepSeek V4 Pro 0813 |
| DTBench | 90.9% | 93.9% | DeepSeek V4 Pro 0813 |
| LMCA | 41.7% | 45.5% | DeepSeek V4 Pro 0813 |
| LMArena Text | Not measured | 1446 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: instruction following | Not measured | 1448 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: creative writing | Not measured | 1439 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: English | Not measured | 1457 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: non-English | Not measured | 1431 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: Chinese | Not measured | 1481 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: Korean | Not measured | 1449 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: French | Not measured | 1446 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: Spanish | Not measured | 1435 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: Russian | Not measured | 1448 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: multi-turn | Not measured | 1454 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: longer queries | Not measured | 1458 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: writing and language | Not measured | 1443 | Only DeepSeek V4 Pro 0813 measured |
| LMArena Text: entertainment and media | Not measured | 1424 | Only DeepSeek V4 Pro 0813 measured |
| Neither model has a result on the other 27 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
DeepSeek V4 Flash 0731 costs less
Repository review
50K input + 3K output tokens
DeepSeek V4 Flash 0731 costs less
Agent loop
200K input + 10K output tokens
DeepSeek V4 Flash 0731 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | DeepSeek V4 Flash 0731 | DeepSeek V4 Pro 0813 |
|---|---|---|
| Maker | DeepSeek | DeepSeek |
| Weights | Open weights | Open weights |
| Released | July 31, 2026 | August 12, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $0.14 | $0.66 |
| Output $/Mtok | $0.28 | $1.98 |
| Coding index | 51.8 | 56.1 |
| Independent measurements behind it | 4 | 5 |
| Cost per solved task | $0.01 | $0.10 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is DeepSeek V4 Flash 0731 better than DeepSeek V4 Pro 0813 for coding?
Pick DeepSeek V4 Pro 0813 for coding: it scores 56.1 on the Vibe Coding Index against 51.8 for DeepSeek V4 Flash 0731. Pick DeepSeek V4 Flash 0731 for output tokens 7.1x cheaper ($0.28 vs $1.98 per million) and a lower cost per solved task ($0.01 against $0.10).
Which is cheaper, DeepSeek V4 Flash 0731 or DeepSeek V4 Pro 0813?
DeepSeek V4 Flash 0731 costs $0.14 per million input tokens and $0.28 per million output tokens; DeepSeek V4 Pro 0813 costs $0.66 and $1.98. Per solved SWE-bench Verified task, DeepSeek V4 Flash 0731 costs $0.01 and DeepSeek V4 Pro 0813 $0.10.
Which has the bigger context window, DeepSeek V4 Flash 0731 or DeepSeek V4 Pro 0813?
Neither: both take 1.05M tokens.
How many benchmarks is the DeepSeek V4 Flash 0731 vs DeepSeek V4 Pro 0813 comparison based on?
45 benchmarks have a result for both models as of October 21, 2026. DeepSeek V4 Flash 0731's coding index rests on 4 independent benchmarks and DeepSeek V4 Pro 0813's on 5. Where only one model has a result, the page says so instead of naming a winner.