Compare / DeepSeek-V4.1-Flash vs Kimi K3

DeepSeek-V4.1-Flash vs Kimi K3

Which is cheaper, which is more capable, and which does better at each kind of work, from 49 results on 3 sources that measured both models.

DeepSeekDeepSeek-V4.1-Flash Released 10 Sep 2026 · $0.30 in · $1.20 out
Moonshot AIKimi K3 Released 16 Jul 2026 · $3.00 in · $15.00 out
DeepSeek-V4.1-Flash wins 2017 too close to callKimi K3 wins 12

Of the 49 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). See also DeepSeek vs Moonshot AI, every model from both labs.

DeepSeek-V4.1-Flash or Kimi K3? The short answer

They score about the same, so pick DeepSeek-V4.1-Flash: it costs 7.7 times less per product listing on CatalogBench.

  • Spring Prompt overall: DeepSeek-V4.1-Flash 51, Kimi K3 47 (out of 100; how the score works).
  • DeepSeek-V4.1-Flash is cheaper to run: $0.006 against $0.047 per product listing on CatalogBench (7.7 times less).
  • Kimi K3 scores higher on the Artificial Analysis Intelligence Index: 43.6 against 39.5.
  • DeepSeek-V4.1-Flash writes faster: 369 tokens a second against 41.
  • DeepSeek-V4.1-Flash is newer: released 10 Sep 2026, against 16 Jul 2026 for Kimi K3.
  • On our own benchmarks, on the results both models have: DeepSeek-V4.1-Flash does better for product listings.
  • On third-party benchmarks, on the results both models have: DeepSeek-V4.1-Flash does better for agents and tool use; Kimi K3 for reasoning and knowledge and writing.

Head to head

51
Spring Prompt overall
47
39.5
Intelligence Index
43.6
$0.52
Price per million tokens, blendedlower is better
$6.00
369
Output speed, tokens a second
41
1,048,576
Context window, tokens
1,048,576

Every fact side by side

DeepSeek-V4.1-FlashKimi K3
Price per million tokens$0.30 in · $1.20 out$3.00 in · $15.00 out
Blended (3 in : 1 out)$0.52$6.00
Cheapest host, blended$0.12 (Morph, FP8)$2.47 (Relace, FP4)
Cost per product listing (CatalogBench)$0.006$0.047
Cached input, per million$0.006 cached input$0.30 cached input
Intelligence Index39.543.6
Spring Prompt overall5147
Output speed, tokens a second36941
Context window1,048,576 tokens1,048,576 tokens
Released10 Sep 2026 (newer)16 Jul 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.

Product listingsTurning a sparse product feed and photos into listings that can go live 15 of 32 21 of 32
Decks from an analysisTurning a finished analysis into a deck you could present as it is not measured 14 of 32
Marketing planningPlanning a year of ad spend without overspending not measured 17 of 32
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 31 of 159 40 of 159
Professional workReal tasks from banking, consulting and law, business documents and freelance projects not measured 24 of 51
Reasoning and knowledgeHard questions across science, maths and general knowledge 42 of 187 18 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 29 of 169 12 of 169
Sticking to the factsSummarising without inventing things, and factual answers 23 of 139 27 of 139
Speed under pressureGood decisions against a real clock (fast chess) not measured 33 of 35

From our research

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-V4.1-FlashKimi K3
CatalogBench measured by Spring Prompt · 3 results
Cost per productUS dollars, lower is better 11 · 31 of 33 measured $0.0061 $0.0469
Reliably publish-ready% of products, higher is better 17 · 17 of 32 measured 41.1% 41.1%
Reliably publish-ready · sales brief% of products, higher is better 11 · 25 of 32 measured 23.2% 1.8%
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 32 · 13 of 387 measured 1,474max reasoning 1,488max reasoning
Overall · Text factualityArena rating, higher is better 28 · 20 of 163 measured 1,467max reasoning 1,472max reasoning
Business, management and finance · TextArena rating, higher is better 22 · 12 of 380 measured 1,479max reasoning 1,487max reasoning
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 32 · 19 of 249 measured 39.5max reasoning 43.6max reasoning
Humanity's Last Exam% of questions, higher is better 50 · 20 of 250 measured 39.2%max reasoning 46.9%max reasoning
Terminal-Bench 4.0% of tasks, higher is better 22 · 37 of 120 measured 26.8%max reasoning 12.6%low reasoning
Show all 49 results (40 more)
MeasureDeepSeek-V4.1-FlashKimi K3
CatalogBench measured by Spring Prompt · 28 results
Channel compliance% of products, higher is better 25 · 25 of 33 measured 99.4% 99.4%
Channel rules broken% of products, lower is better 32 · 20 of 33 measured 3.0% 0.6%
Channel rules broken · sales brief% of products, lower is better 29 · 17 of 33 measured 1.8% 0.6%
Claims to check% of products, lower is better 1 · 1 of 31 measured 0.0% 0.0%
Claims to check · sales brief% of products, lower is better 1 · 1 of 31 measured 0.0% 0.0%
Conflicts caught% of conflicts, higher is better 1 · 31 of 33 measured 100.0% 97.0%
Content quality% of checks, higher is better 20 · 23 of 33 measured 96.7% 96.4%
Content quality · sales brief% of checks, higher is better 23 · 20 of 33 measured 96.2% 96.4%
Decision accuracy% of decisions, higher is better 15 · 20 of 33 measured 96.9% 96.1%
Failed outputs% of products, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Failed outputs · sales brief% of products, lower is better 26 · 26 of 33 measured 0.6% 0.6%
Field accuracy% of missing fields, higher is better 8 · 8 of 33 measured 94.4% 94.4%
Invented values% of filled values, lower is better 28 · 32 of 33 measured 3.1% 5.7%
Missing UK information% of products, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Missing UK information · sales brief% of products, lower is better 1 · 1 of 33 measured 0.0% 0.0%
Not findable% of products, lower is better 12 · 17 of 33 measured 3.6% 4.2%
Not findable · sales brief% of products, lower is better 16 · 14 of 33 measured 3.6% 3.0%
Publish-ready listings% of products, higher is better 16 · 18 of 33 measured 63.7% 61.3%
Publish-ready listings · sales brief% of products, higher is better 12 · 24 of 33 measured 47.0% 14.9%
Time per productseconds, lower is better 18 · 31 of 33 measured 17.6 66.6
Unsupported claimsclaims per product, lower is better 17 · 24 of 33 measured 0.54 0.98
Unsupported claims% of products, lower is better 13 · 16 of 33 measured 13.1% 16.7%
Unsupported claims · sales briefclaims per product, lower is better 12 · 27 of 33 measured 1.25 3.64
Unsupported claims · sales brief% of products, lower is better 13 · 28 of 33 measured 39.9% 81.6%
Wrong attributes% of products, lower is better 26 · 30 of 33 measured 20.2% 24.4%
Wrong attributes · sales brief% of products, lower is better 23 · 25 of 33 measured 20.2% 20.8%
Wrong category or variant% of products, lower is better 24 · 24 of 33 measured 3.0% 3.0%
Wrong category or variant · sales brief% of products, lower is better 18 · 28 of 33 measured 2.4% 4.2%
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 8 results
Confirmed task success · AgentIPS effect estimate, higher is better 8 · 9 of 48 measured 0.08max reasoning, arena agent 0.08max reasoning, arena agent
Creative writing · TextArena rating, higher is better 49 · 20 of 385 measured 1,440max reasoning 1,460max reasoning
Expert prompts · TextArena rating, higher is better 18 · 10 of 338 measured 1,517max reasoning 1,533max reasoning
Instruction following · TextArena rating, higher is better 18 · 11 of 387 measured 1,479max reasoning 1,488max reasoning
Praise over complaint · AgentIPS effect estimate, higher is better 23 · 16 of 48 measured 0.02max reasoning, arena agent 0.07max reasoning, arena agent
Steerability · AgentIPS effect estimate, higher is better 24 · 23 of 48 measured 0.01max reasoning, arena agent 0.01max reasoning, arena agent
Tool grounding · AgentIPS effect estimate, higher is better 16 · 2 of 48 measured 0.00max reasoning, arena agent 0.00max reasoning, arena agent
Writing, literature and language · TextArena rating, higher is better 37 · 13 of 386 measured 1,454max reasoning 1,476max reasoning
Artificial Analysis reported by Artificial Analysis · 4 results
Long-context reasoning (AA-LCR)% of questions, higher is better 7 · 1 of 241 measured 84.0%max reasoning 88.7%max reasoning
Output speedtokens per second, higher is better 3 · 74 of 78 measured 369no reasoning 41.0max reasoning
SciCode% of problems, higher is better 44 · 11 of 121 measured 51.9%max reasoning 59.5%max reasoning
Time to first answer tokenseconds, lower is better 13 · 76 of 78 measured 0.63no reasoning 51.8max reasoning

Quick answers

Is DeepSeek-V4.1-Flash better than Kimi K3?

They score about the same, so pick DeepSeek-V4.1-Flash: it costs 7.7 times less per product listing on CatalogBench.

Which is cheaper, DeepSeek-V4.1-Flash or Kimi K3?

DeepSeek-V4.1-Flash is cheaper to run: $0.006 against $0.047 per product listing on CatalogBench (7.7 times less).

Which is faster, DeepSeek-V4.1-Flash or Kimi K3?

DeepSeek-V4.1-Flash writes about 369 tokens a second on its usual API, against 41 for Kimi K3 (Artificial Analysis).

Which is better for product listings, DeepSeek-V4.1-Flash or Kimi K3?

DeepSeek-V4.1-Flash ranks 15 of 32 models on sale for turning a sparse product feed and photos into listings that can go live, against 21 for Kimi K3.

Which is better for agents and tool use, DeepSeek-V4.1-Flash or Kimi K3?

DeepSeek-V4.1-Flash ranks 31 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 40 for Kimi K3.

Which is better for reasoning and knowledge, DeepSeek-V4.1-Flash or Kimi K3?

Kimi K3 ranks 18 of 187 models on sale for hard questions across science, maths and general knowledge, against 42 for DeepSeek-V4.1-Flash.

Which is better for writing, DeepSeek-V4.1-Flash or Kimi K3?

Kimi K3 ranks 12 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 29 for DeepSeek-V4.1-Flash.

Run DeepSeek-V4.1-Flash and Kimi K3 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.