Agentic
Like-for-like- DeepSeek V4 Flash 0731
- 63.8
- Kimi K3
- 89.5
- Weighted basis
- 2 vs 2 rows
- Reading
- Kimi K3 leads
DeepSeek V4 Flash 0731 vs Kimi K3
Updated August 13, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
9 results are shared. Category rows based on different benchmark sets are marked directional and do not name a winner.
Which one for your work
Recommendations appear only when a shared evidence basis or an explicit operating constraint supports the call. Secondary and unsupported use cases stay disclosed below the initial list.
Agentic work
Tool use, computer use, and multi-step task completion
Kimi K3
Kimi K3 leads on the same 2 weighted benchmark rows.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Kimi K3
Kimi K3 has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Show secondary and unsupported calls
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Repository review cost
50K fresh input + 3K output tokens
DeepSeek V4 Flash 0731
DeepSeek V4 Flash 0731 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Coding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
No shared weighted benchmark basis supports a winner.
Confidence: limited
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
24 DeepSeek V4 Flash 0731 only
9 Shared
31 Kimi K3 only
- Shared results
- 9
- DeepSeek V4 Flash 0731 only
- 24
- Kimi K3 only
- 31
- Like-for-like categories
- 1 / 8
1 category uses different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row states whether both averages use the same weighted benchmark set. Directional and not-comparable rows remain visible, but they never receive a winner in this template.
Knowledge
Directional only- DeepSeek V4 Flash 0731
- 55.3
- Kimi K3
- 61.0
- Weighted basis
- 4 vs 2 rows
- Reading
- Directional only
Coding
Not comparable- DeepSeek V4 Flash 0731
- 68.8
- Kimi K3
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Reasoning
Not comparable- DeepSeek V4 Flash 0731
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- DeepSeek V4 Flash 0731
- 94.8
- Kimi K3
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- DeepSeek V4 Flash 0731
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- DeepSeek V4 Flash 0731
- Not measured
- Kimi K3
- 78.5
- Weighted basis
- 0 vs 3 rows
- Reading
- Not comparable
Instruction following
Not comparable- DeepSeek V4 Flash 0731
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | DeepSeek V4 Flash 0731 | Kimi K3 | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | 63.8 | 89.5 | Like-for-like2 vs 2 rows | Kimi K3 leads |
| Knowledge | 55.3 | 61.0 | Directional only4 vs 2 rows | Directional only |
| Coding | 68.8 | Not measured | Not comparable2 vs 0 rows | Not comparable |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Math | 94.8 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 78.5 | Not comparable0 vs 3 rows | Not comparable |
| Instruction following | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
Shape of the matched evidence
Only shared public evidence is shown. Sparse evidence stays a ruled list rather than being closed into a radar shape.
Too few matched category axes support a radar. The ruled list below shows only shared benchmark results; positions use each benchmark’s normalized display scale when available.
Terminal-Bench 2.0
Agentic
DeepSeek V4 Flash 0731: 56.9%Kimi K3: 88.3%Normalized gap 31.4DeepSeek V4 Flash 0731 source Kimi K3 sourceHLE
Knowledge
DeepSeek V4 Flash 0731: 34.8%Kimi K3: 56%Normalized gap 21.2DeepSeek V4 Flash 0731 source Kimi K3 sourceBrowseComp
Agentic
DeepSeek V4 Flash 0731: 73.2%Kimi K3: 91.2%Normalized gap 18.0DeepSeek V4 Flash 0731 source Kimi K3 sourceGPQA
Knowledge
DeepSeek V4 Flash 0731: 88.1%Kimi K3: 93.5%Normalized gap 5.4DeepSeek V4 Flash 0731 source Kimi K3 source
What each workload costs
Three fixed token mixes turn per-token rates into comparable decisions. Each scenario states context fit and whether cached input had to fall back to the published list-input rate.
Chat turn
1K fresh input + 500 output tokens
- DeepSeek V4 Flash 0731
- $0.00028
- Fits in one request
- Kimi K3
- $0.0105
- Fits in one request
DeepSeek V4 Flash 0731 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- DeepSeek V4 Flash 0731
- $0.00784
- Fits in one request
- Kimi K3
- $0.195
- Fits in one request
DeepSeek V4 Flash 0731 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- DeepSeek V4 Flash 0731
- $0.00616
- Fits in one request
- Kimi K3
- $0.27
- Fits in one request
DeepSeek V4 Flash 0731 has the lower modeled cost
Costs use the listed standard API rates.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationDeepSeek V4 Flash 0731Kimi K3
Context window
Maximum documented context; output-token limits may be lower.
DeepSeek V4 Flash 0731
Kimi K3
1.05M
API model ID
DeepSeek V4 Flash 0731
deepseek-v4-flash
DeepSeek V4 Flash 0731 updateKimi K3
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
DeepSeek V4 Flash 0731
$0.0028 per 1M cached input tokens
Kimi K3
$0.3 per 1M cached input tokens
Documented inputs
DeepSeek V4 Flash 0731
Kimi K3
Not sourced
Documented outputs
DeepSeek V4 Flash 0731
Kimi K3
Not sourced
Provider availability
DeepSeek V4 Flash 0731
Public Beta · DeepSeek API
DeepSeek V4 Flash 0731 updateKimi K3
Not sourced
Reasoning profile
DeepSeek V4 Flash 0731
Reasoning
Kimi K3
Reasoning
Weight access
DeepSeek V4 Flash 0731
Proprietary
Kimi K3
Pending
License
DeepSeek V4 Flash 0731
Proprietary
Kimi K3
Pending
Release date
DeepSeek V4 Flash 0731
2026-07-31
Kimi K3
2026-07-16
If you already use one of these models
- Deployment change
- The models list different providers, so authentication, endpoint behavior, limits, and feature support may change.
- Quality signal
- At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner.
- Workload cost
- Repository review: $0.00784 vs $0.195. Cache-heavy agent loop: $0.00616 vs $0.27.
- Context tradeoff
- Kimi K3 has the larger documented window (1.05M).
Run the same representative tasks against both endpoints before changing production traffic.
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence64 rows
Agentic
Terminal-Bench 2.0
DeepSeek V4 Flash 073156.9%SourceKimi K388.3%SourceKimi K3 leads this result
BrowseComp
DeepSeek V4 Flash 073173.2%SourceKimi K391.2%SourceKimi K3 leads this result
HLE w/ tools
DeepSeek V4 Flash 073145.1%SourceKimi K3—Not directly comparable
MCP Atlas
DeepSeek V4 Flash 073169%SourceKimi K384.2%SourceKimi K3 leads this result
Toolathlon
DeepSeek V4 Flash 073147.8%SourceKimi K3—Not directly comparable
Terminal-Bench 2.1
DeepSeek V4 Flash 073182.7%SourceKimi K3—Not directly comparable
CyberGym
DeepSeek V4 Flash 073176.7%SourceKimi K3—Not directly comparable
Toolathlon-Verified
DeepSeek V4 Flash 073170.3%SourceKimi K373.2%SourceKimi K3 leads this result
Agents' Last Exam
DeepSeek V4 Flash 073125.2%SourceKimi K3—Not directly comparable
AutomationBench
DeepSeek V4 Flash 073125.1%SourceKimi K330.8%SourceKimi K3 leads this result
DeepSearchQA
DeepSeek V4 Flash 0731—Kimi K395.0%SourceNot directly comparable
JobBench
DeepSeek V4 Flash 0731—Kimi K352.9%SourceNot directly comparable
APEX-Agents
DeepSeek V4 Flash 0731—Kimi K337.6%SourceNot directly comparable
SpreadsheetBench 2
DeepSeek V4 Flash 0731—Kimi K334.8%SourceNot directly comparable
DECK-Bench
DeepSeek V4 Flash 0731—Kimi K373.5%SourceNot directly comparable
Coding
LiveCodeBench Pass@1-COT
DeepSeek V4 Flash 073191.6%SourceKimi K3—Not directly comparable
Codeforces
DeepSeek V4 Flash 07313052.0SourceKimi K3—Not directly comparable
SWE-bench Verified
DeepSeek V4 Flash 073179%SourceKimi K3—Not directly comparable
SWE-bench Pro
DeepSeek V4 Flash 073152.6%SourceKimi K3—Not directly comparable
SWE Multilingual
DeepSeek V4 Flash 073173.3%SourceKimi K3—Not directly comparable
Terminal-Bench 2.0
DeepSeek V4 Flash 073156.9%SourceKimi K3—Not directly comparable
Terminal-Bench 2.1
DeepSeek V4 Flash 073182.7%SourceKimi K3—Not directly comparable
NL2Repo
DeepSeek V4 Flash 073154.2%SourceKimi K3—Not directly comparable
deepSwe
DeepSeek V4 Flash 073154.4%SourceKimi K367.5%SourceKimi K3 leads this result
DSBench-FullStack
DeepSeek V4 Flash 073168.7%SourceKimi K3—Not directly comparable
DSBench-Hard
DeepSeek V4 Flash 073159.6%SourceKimi K3—Not directly comparable
cursorBench32
DeepSeek V4 Flash 0731—Kimi K360.8%SourceNot directly comparable
FrontierSWE
DeepSeek V4 Flash 0731—Kimi K381.2%SourceNot directly comparable
ProgramBench
DeepSeek V4 Flash 0731—Kimi K377.8%SourceNot directly comparable
Kimi Code Bench v2
DeepSeek V4 Flash 0731—Kimi K372.9%SourceNot directly comparable
sweMarathon
DeepSeek V4 Flash 0731—Kimi K342%SourceNot directly comparable
PostTrain Bench
DeepSeek V4 Flash 0731—Kimi K336.6%SourceNot directly comparable
MLS-Bench Lite
DeepSeek V4 Flash 0731—Kimi K348.3%SourceNot directly comparable
VulcanBench v3
DeepSeek V4 Flash 0731—Kimi K373.9%SourceNot directly comparable
APEX-SWE
DeepSeek V4 Flash 0731—Kimi K348.0%SourceNot directly comparable
EEBench
DeepSeek V4 Flash 0731—Kimi K338.3%SourceNot directly comparable
InferenceEval
DeepSeek V4 Flash 0731—Kimi K339.8%SourceNot directly comparable
KernelBench Internal
DeepSeek V4 Flash 0731—Kimi K368.5%SourceNot directly comparable
Reasoning
Knowledge
MMLU-Pro
DeepSeek V4 Flash 073186.2%SourceKimi K3—Not directly comparable
SimpleQA
DeepSeek V4 Flash 073134.1%SourceKimi K3—Not directly comparable
Chinese-SimpleQA
DeepSeek V4 Flash 073178.9%SourceKimi K3—Not directly comparable
GPQA
DeepSeek V4 Flash 073188.1%SourceKimi K393.5%SourceKimi K3 leads this result
GPQA-D
DeepSeek V4 Flash 073188.1%SourceKimi K393.5%SourceKimi K3 leads this result
HLE
DeepSeek V4 Flash 073134.8%SourceKimi K356%SourceKimi K3 leads this result
HLE w/o tools
DeepSeek V4 Flash 0731—Kimi K343.5%SourceNot directly comparable
Math
HMMT Feb 2026
DeepSeek V4 Flash 073194.8%SourceKimi K3—Not directly comparable
IMOAnswerBench
DeepSeek V4 Flash 073188.4%SourceKimi K3—Not directly comparable
Apex
DeepSeek V4 Flash 073133.0%SourceKimi K3—Not directly comparable
Apex Shortlist
DeepSeek V4 Flash 073185.7%SourceKimi K3—Not directly comparable
Multimodal
OfficeQA Pro
DeepSeek V4 Flash 0731—Kimi K363.3%SourceNot directly comparable
MMMU-Pro
DeepSeek V4 Flash 0731—Kimi K381.6%SourceNot directly comparable
MMMU-Pro w/ Python
DeepSeek V4 Flash 0731—Kimi K383.4%SourceNot directly comparable
CharXiv w/o tools
DeepSeek V4 Flash 0731—Kimi K384.8%SourceNot directly comparable
CharXiv
DeepSeek V4 Flash 0731—Kimi K391.3%SourceNot directly comparable
MathVision
DeepSeek V4 Flash 0731—Kimi K394.3%SourceNot directly comparable
MathVision w/ Python
DeepSeek V4 Flash 0731—Kimi K397.8%SourceNot directly comparable
BabyVision w/ Python
DeepSeek V4 Flash 0731—Kimi K385.7%SourceNot directly comparable
ZeroBench
DeepSeek V4 Flash 0731—Kimi K323.0%SourceNot directly comparable
ZeroBench w/ Python
DeepSeek V4 Flash 0731—Kimi K341.0%SourceNot directly comparable
WorldVQA ForceAnswer
DeepSeek V4 Flash 0731—Kimi K351.0%SourceNot directly comparable
OmniDocBench
DeepSeek V4 Flash 0731—Kimi K391.1%SourceNot directly comparable
PerceptionBench
DeepSeek V4 Flash 0731—Kimi K358.5%SourceNot directly comparable
Frequently asked questions
Which is better, DeepSeek V4 Flash 0731 or Kimi K3?
At least one model is not scored in the current public ranking lane, so the page does not name an overall quality winner. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, DeepSeek V4 Flash 0731 or Kimi K3?
The published evidence does not provide a shared weighted coding basis for both models, so BenchLM does not name a coding winner.
Which is better for agentic tasks, DeepSeek V4 Flash 0731 or Kimi K3?
Kimi K3 leads the like-for-like agentic tasks comparison across 2 shared weighted benchmark rows.
Which costs less, DeepSeek V4 Flash 0731 or Kimi K3?
For the stated presets, chat costs $0.00028 on DeepSeek V4 Flash 0731 and $0.0105 on Kimi K3; repository review costs $0.00784 and $0.195; the cache-heavy agent loop costs $0.00616 and $0.27. Costs use the listed standard API rates.
Which has the larger context window, DeepSeek V4 Flash 0731 or Kimi K3?
Kimi K3 has the larger documented context window: 1.05M, compared with 1M.
Related comparisons
Last updated August 13, 2026
Watch DeepSeek V4 Flash 0731 vs Kimi K3
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.