DeepSeek V4.1 Flash vs Step 5 Preview
DeepSeek V4.1 Flash leads Step 5 Preview on coding, 62.5 to 61.1 on the Vibe Coding Index, and DeepSeek V4.1 Flash costs 3.2x less per output token ($0.84 vs $2.70 per million). Updated October 22, 2026; 61 benchmarks have a result for both.
DeepSeek V4.1 Flash or Step 5 Preview: which to pick
DeepSeek V4.1 Flash and Step 5 Preview are level for coding, 62.5 against 61.1 on the Vibe Coding Index, so price and context decide.
Pick Step 5 Preview for work measured by Terminal-Bench 4 (Vals AI) or SciCode, where it leads. Pick DeepSeek V4.1 Flash for output tokens 3.2x cheaper ($0.84 vs $2.70 per million) and work measured by LiveBench coding or LMArena WebDev, where it leads.
- Shared results
- 61
- DeepSeek V4.1 Flash only
- 40
- Step 5 Preview only
- 0
- Like-for-like categories
- 8 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
No call
Level: 62.5 against 61.1, inside 2 points.
5 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
Step 5 Preview
Step 5 Preview leads, 43.7 against 24.7.
Same board, both measured
Web design
Design Arena: websites
No call
Only DeepSeek V4.1 Flash has a Design Arena: websites result, 1308.
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 V4.1 Flash
DeepSeek V4.1 Flash $0.0006, Step 5 Preview $0.0024 per turn.
List prices
At a glance
- Coding index
- DeepSeek V4.1 Flash 62.5 from 5 benchmarks, Step 5 Preview 61.1 from 4 benchmarks.
- Widest category
- Intelligence, 19.0 points apart on average over 1 shared benchmark.
- Widest single gap
- LiveBench browser: DeepSeek V4.1 Flash 84.0% against Step 5 Preview 24.0%, 60.0 points apart on a 0 to 100 scale.
- Output price
- DeepSeek V4.1 Flash $0.84 against Step 5 Preview $2.70 per million output tokens, 3.2x 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 V4.1 Flash | Bars | Step 5 Preview | Ties |
|---|---|---|---|---|
| Coding | 11 | 2 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 9 | 2 | ||
| Reasoning | 6 | 2 | ||
| Math | 3 | 0 | ||
| Knowledge | 4 | 0 | ||
| Web design | 1 | 0 | ||
| General | 18 | 2 | ||
| All shared boards | 52 | 9 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index112
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench coding | 90.2% | 85.6% | DeepSeek V4.1 Flash |
| LMArena WebDev | 1618 | 1567 | DeepSeek V4.1 Flash |
| Terminal-Bench 4 (Vals AI) | 19.7% | 31.8% | Step 5 Preview |
| Vibe Code Bench (Vals AI) | 84.7% | 70.4% | DeepSeek V4.1 Flash |
| Code migration (Vals AI) | 45.6% | 39.0% | DeepSeek V4.1 Flash |
| ProgramBench (Vals AI) | 0.5% | Not measured | Only DeepSeek V4.1 Flash measured |
| IOI (Vals AI) | 40.3% | 38.1% | DeepSeek V4.1 Flash |
| Terminal-Bench 2.1 (Vals AI) | 74.5% | Not measured | Only DeepSeek V4.1 Flash measured |
| SWE-bench Multimodal (Vals AI) | 46.5% | 40.8% | DeepSeek V4.1 Flash |
| Vibe Code Bench 1-100 (Vals AI) | 16.4% | Not measured | Only DeepSeek V4.1 Flash measured |
| SkillsBench (Vals AI) | 69.8% | Not measured | Only DeepSeek V4.1 Flash measured |
| CyberBench (Vals AI) | 73.7% | Not measured | Only DeepSeek V4.1 Flash measured |
| SciCode | 51.9% | 58.9% | Step 5 Preview |
| FrontierCode | 43.9% | Not measured | Only DeepSeek V4.1 Flash measured |
| ALE-Bench | 1092.3 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: coding | 1502 | 1485 | DeepSeek V4.1 Flash |
| LMArena Text: software and IT | 1496 | 1473 | DeepSeek V4.1 Flash |
| LMArena WebDev: React | 1628 | 1583 | DeepSeek V4.1 Flash |
| LMArena WebDev: HTML | 1612 | 1556 | DeepSeek V4.1 Flash |
| LMArena WebDev: from an image | 1606 | 1550 | DeepSeek V4.1 Flash |
| Neither model has a result on the other 19 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 24.7 | 43.7 | Step 5 Preview |
| Epoch Capabilities Index | 155.9 | Not measured | Only DeepSeek V4.1 Flash measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index92
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench research | 86.8% | 72.8% | DeepSeek V4.1 Flash |
| LiveBench browser | 84.0% | 24.0% | DeepSeek V4.1 Flash |
| LiveBench long agentic | 40.1% | 27.5% | DeepSeek V4.1 Flash |
| SREBench (Vals AI) | 0.8% | 0.4% | DeepSeek V4.1 Flash |
| Finance Agent (Vals AI) | 53.5% | 50.7% | DeepSeek V4.1 Flash |
| Tax Agent Bench (Vals AI) | 26.6% | 27.3% | Step 5 Preview |
| Legal Research Bench (Vals AI) | 41.4% | 40.4% | DeepSeek V4.1 Flash |
| Excel Modeling Benchmark (Vals AI) | 57.2% | 57.1% | DeepSeek V4.1 Flash |
| BioMysteryBench (Vals AI) | 67.8% | 67.4% | DeepSeek V4.1 Flash |
| Terminal-Bench-Science (Vals AI) | 0.0% | 2.9% | Step 5 Preview |
| Vals Science Index | 29.6% | 29.6% | DeepSeek V4.1 Flash |
| APEX-Agents | 39.5% | Not measured | Only DeepSeek V4.1 Flash measured |
| Neither model has a result on the other 29 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index62
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench reasoning | 86.7% | 83.8% | DeepSeek V4.1 Flash |
| MysteryMechanism (Vals AI) | 21.2% | 18.5% | DeepSeek V4.1 Flash |
| Humanity's Last Exam (Artificial Analysis run) | 39.2% | 46.5% | Step 5 Preview |
| GPQA Diamond (Epoch AI run) | 89.8% | Not measured | Only DeepSeek V4.1 Flash measured |
| ARC-AGI | 94.5% | Not measured | Only DeepSeek V4.1 Flash measured |
| ARC-AGI-2 | 72.9% | Not measured | Only DeepSeek V4.1 Flash measured |
| Chess puzzles | 48.0% | Not measured | Only DeepSeek V4.1 Flash measured |
| Mystery game puzzles | 43.0% | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: hard prompts | 1483 | 1461 | DeepSeek V4.1 Flash |
| LMArena Text: expert | 1501 | 1485 | DeepSeek V4.1 Flash |
| LMArena Text: hard prompts in English | 1488 | 1454 | DeepSeek V4.1 Flash |
| LMArena Vision: diagrams | 1297 | 1308 | Step 5 Preview |
| LMArena Vision: homework | 1307 | 1265 | DeepSeek V4.1 Flash |
| Neither model has a result on the other 9 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| OTIS Mock AIME | 98.3% | Not measured | Only DeepSeek V4.1 Flash measured |
| FrontierMath, tiers 1 to 3 | 67.4% | Not measured | Only DeepSeek V4.1 Flash measured |
| FrontierMath, tier 4 | 26.8% | Not measured | Only DeepSeek V4.1 Flash measured |
| ProofBench | 54.0% | 42.0% | DeepSeek V4.1 Flash |
| LMArena Text: math | 1478 | 1458 | DeepSeek V4.1 Flash |
| LMArena Text: mathematical work | 1485 | 1460 | DeepSeek V4.1 Flash |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | 83.3% | Not measured | Only DeepSeek V4.1 Flash measured |
| MedCode (Vals AI) | 41.2% | Not measured | Only DeepSeek V4.1 Flash measured |
| SimpleQA Verified | 42.9% | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: science | 1469 | 1454 | DeepSeek V4.1 Flash |
| LMArena Text: medicine and healthcare | 1484 | 1439 | DeepSeek V4.1 Flash |
| LMArena Text: legal and government | 1443 | 1437 | DeepSeek V4.1 Flash |
| LMArena Text: business and finance | 1461 | 1435 | DeepSeek V4.1 Flash |
| Neither model has a result on the other 11 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1308 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: UI components | 1312 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: game dev | 1330 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: data viz | 1291 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: 3D | 1355 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: all code categories | 1320 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: SVG | 1245 | Not measured | Only DeepSeek V4.1 Flash measured |
| Design Arena: ASCII art | 1265 | Not measured | Only DeepSeek V4.1 Flash measured |
| LiveBench design | 79.2% | 76.9% | DeepSeek V4.1 Flash |
GeneralLike-for-likePublic boards, shown as published, kept out of the index182
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench data analysis | 79.3% | 63.5% | DeepSeek V4.1 Flash |
| Public Benefits Bench (Vals AI) | 64.3% | 58.1% | DeepSeek V4.1 Flash |
| SAGE (Vals AI) | 47.9% | Not measured | Only DeepSeek V4.1 Flash measured |
| MedScribe (Vals AI) | 85.5% | Not measured | Only DeepSeek V4.1 Flash measured |
| Long context reasoning (AA-LCR) | 84.0% | 88.3% | Step 5 Preview |
| DTBench | 89.9% | Not measured | Only DeepSeek V4.1 Flash measured |
| LMCA | 47.0% | Not measured | Only DeepSeek V4.1 Flash measured |
| Surface Evolver Bench | 46.3% | Not measured | Only DeepSeek V4.1 Flash measured |
| Furniture assembly | 34.2% | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text | 1464 | 1440 | DeepSeek V4.1 Flash |
| LMArena Text: instruction following | 1473 | 1438 | DeepSeek V4.1 Flash |
| LMArena Text: creative writing | 1437 | 1409 | DeepSeek V4.1 Flash |
| LMArena Vision | 1278 | 1265 | DeepSeek V4.1 Flash |
| LMArena Text: English | 1473 | 1447 | DeepSeek V4.1 Flash |
| LMArena Text: non-English | 1449 | 1428 | DeepSeek V4.1 Flash |
| LMArena Text: Chinese | 1501 | 1500 | DeepSeek V4.1 Flash |
| LMArena Text: German | 1470 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: Japanese | 1432 | 1383 | DeepSeek V4.1 Flash |
| LMArena Text: Korean | 1436 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: French | 1465 | 1438 | DeepSeek V4.1 Flash |
| LMArena Text: Spanish | 1459 | 1441 | DeepSeek V4.1 Flash |
| LMArena Text: Russian | 1467 | 1433 | DeepSeek V4.1 Flash |
| LMArena Text: Polish | 1457 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Text: multi-turn | 1461 | 1437 | DeepSeek V4.1 Flash |
| LMArena Text: longer queries | 1473 | 1449 | DeepSeek V4.1 Flash |
| LMArena Text: writing and language | 1450 | 1425 | DeepSeek V4.1 Flash |
| LMArena Text: entertainment and media | 1445 | 1412 | DeepSeek V4.1 Flash |
| LMArena Vision: English | 1288 | 1290 | Step 5 Preview |
| LMArena Vision: Chinese | 1327 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Vision: reading text | 1282 | 1282 | DeepSeek V4.1 Flash |
| LMArena Vision: humor | 1260 | Not measured | Only DeepSeek V4.1 Flash measured |
| LMArena Vision: creative writing | 1275 | Not measured | Only DeepSeek V4.1 Flash measured |
| Neither model has a result on the other 14 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | DeepSeek V4.1 Flash | Step 5 Preview | 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 V4.1 Flash costs less
Repository review
50K input + 3K output tokens
DeepSeek V4.1 Flash costs less
Agent loop
200K input + 10K output tokens
DeepSeek V4.1 Flash 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 V4.1 Flash | StepFun Step 5 Preview |
|---|---|---|
| Maker | DeepSeek | StepFun |
| Weights | Open weights | Open weights |
| Released | September 10, 2026 | October 8, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $0.2225 | $1.00 |
| Output $/Mtok | $0.84 | $2.70 |
| Coding index | 62.5 | 61.1 |
| Independent measurements behind it | 5 | 4 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is DeepSeek V4.1 Flash better than Step 5 Preview for coding?
DeepSeek V4.1 Flash and Step 5 Preview are level for coding, 62.5 against 61.1 on the Vibe Coding Index, so price and context decide. Pick Step 5 Preview for work measured by Terminal-Bench 4 (Vals AI) or SciCode, where it leads. Pick DeepSeek V4.1 Flash for output tokens 3.2x cheaper ($0.84 vs $2.70 per million) and work measured by LiveBench coding or LMArena WebDev, where it leads.
Which is cheaper, DeepSeek V4.1 Flash or Step 5 Preview?
DeepSeek V4.1 Flash costs $0.2225 per million input tokens and $0.84 per million output tokens; Step 5 Preview costs $1.00 and $2.70. Cost per solved task is known for neither of them.
Which has the bigger context window, DeepSeek V4.1 Flash or Step 5 Preview?
Neither: both take 1.05M tokens.
How many benchmarks is the DeepSeek V4.1 Flash vs Step 5 Preview comparison based on?
61 benchmarks have a result for both models as of October 22, 2026. DeepSeek V4.1 Flash's coding index rests on 5 independent benchmarks and Step 5 Preview's on 4. Where only one model has a result, the page says so instead of naming a winner.