Agentic
Directional only- DeepSeek V4 Pro 0813
- 74.5
- Kimi K2.6
- 73.5
- Weighted basis
- 2 vs 3 rows
- Reading
- Directional only
DeepSeek V4 Pro 0813 vs Kimi K2.6
Updated August 13, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
DeepSeek V4 Pro 0813 has the higher public score estimate, 60.93 versus 60, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
13 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.
Long documents
Prompts that approach the documented context limit
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 has the lower estimated token cost for this stated workload. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Confidence: rate-fallback
Show secondary and unsupported calls
Repository review cost
50K fresh input + 3K output tokens
DeepSeek V4 Pro 0813
DeepSeek V4 Pro 0813 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
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
The category averages use different weighted benchmark sets, so they are directional rather than like-for-like.
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.
21 DeepSeek V4 Pro 0813 only
13 Shared
19 Kimi K2.6 only
- Shared results
- 13
- DeepSeek V4 Pro 0813 only
- 21
- Kimi K2.6 only
- 19
- Like-for-like categories
- 0 / 8
4 categories use 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.
Coding
Directional only- DeepSeek V4 Pro 0813
- 70.9
- Kimi K2.6
- 64.4
- Weighted basis
- 2 vs 3 rows
- Reading
- Directional only
Knowledge
Directional only- DeepSeek V4 Pro 0813
- 62.5
- Kimi K2.6
- 42.2
- Weighted basis
- 4 vs 2 rows
- Reading
- Directional only
Math
Directional only- DeepSeek V4 Pro 0813
- 95.2
- Kimi K2.6
- 67.1
- Weighted basis
- 1 vs 4 rows
- Reading
- Directional only
Reasoning
Not comparable- DeepSeek V4 Pro 0813
- Not measured
- Kimi K2.6
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- DeepSeek V4 Pro 0813
- Not measured
- Kimi K2.6
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- DeepSeek V4 Pro 0813
- Not measured
- Kimi K2.6
- 79.8
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Instruction following
Not comparable- DeepSeek V4 Pro 0813
- Not measured
- Kimi K2.6
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | DeepSeek V4 Pro 0813 | Kimi K2.6 | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | 74.5 | 73.5 | Directional only2 vs 3 rows | Directional only |
| Coding | 70.9 | 64.4 | Directional only2 vs 3 rows | Directional only |
| Knowledge | 62.5 | 42.2 | Directional only4 vs 2 rows | Directional only |
| Math | 95.2 | 67.1 | Directional only1 vs 4 rows | Directional only |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Multimodal | Not measured | 79.8 | Not comparable0 vs 2 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.
HLE
Knowledge
DeepSeek V4 Pro 0813: 42.7%Kimi K2.6: 34.7%Normalized gap 8.0DeepSeek V4 Pro 0813 source Kimi K2.6 sourceSWE-bench Pro
Coding
DeepSeek V4 Pro 0813: 55.4%Kimi K2.6: 58.6%Normalized gap 3.2DeepSeek V4 Pro 0813 source Kimi K2.6 sourceHMMT Feb 2026
Math
DeepSeek V4 Pro 0813: 95.2%Kimi K2.6: 92.7%Normalized gap 2.5DeepSeek V4 Pro 0813 source Kimi K2.6 sourceTerminal-Bench 2.0
Agentic
DeepSeek V4 Pro 0813: 67.9%Kimi K2.6: 66.7%Normalized gap 1.2DeepSeek V4 Pro 0813 source Kimi K2.6 sourceGPQA
Knowledge
DeepSeek V4 Pro 0813: 90.1%Kimi K2.6: 90.5%Normalized gap 0.4DeepSeek V4 Pro 0813 source Kimi K2.6 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 Pro 0813
- $0.00087
- Fits in one request
- Kimi K2.6
- $0.00295
- Fits in one request
DeepSeek V4 Pro 0813 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- DeepSeek V4 Pro 0813
- $0.02436
- Fits in one request
- Kimi K2.6
- $0.0595
- Fits in one request
DeepSeek V4 Pro 0813 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 Pro 0813
- $0.01812
- Fits in one request
- Kimi K2.6
- $0.249
- Fits in one request
- Cached input priced at the published list-input rate
DeepSeek V4 Pro 0813 has the lower modeled cost
Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationDeepSeek V4 Pro 0813Kimi K2.6
Context window
Maximum documented context; output-token limits may be lower.
DeepSeek V4 Pro 0813
Kimi K2.6
256K
API model ID
DeepSeek V4 Pro 0813
deepseek-v4-pro
DeepSeek models and pricingKimi K2.6
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 Pro 0813
$0.003625 per 1M cached input tokens
Kimi K2.6
Not published
Documented inputs
DeepSeek V4 Pro 0813
Kimi K2.6
Not sourced
Documented outputs
DeepSeek V4 Pro 0813
Kimi K2.6
Not sourced
Provider availability
DeepSeek V4 Pro 0813
Generally Available · DeepSeek API
DeepSeek models and pricingKimi K2.6
Not sourced
Reasoning profile
DeepSeek V4 Pro 0813
Reasoning
Kimi K2.6
Reasoning
Weight access
DeepSeek V4 Pro 0813
Proprietary
Kimi K2.6
Open Weight
License
DeepSeek V4 Pro 0813
Proprietary
Kimi K2.6
Open Weight
Release date
DeepSeek V4 Pro 0813
2026-08-13
Kimi K2.6
2026-04-20
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
- DeepSeek V4 Pro 0813 has the higher public score estimate, 60.93 versus 60, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.02436 vs $0.0595. Cache-heavy agent loop: $0.01812 vs $0.249.
- Context tradeoff
- DeepSeek V4 Pro 0813 has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
DeepSeek V4 Pro 0813
API / mo$0
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/day
Benchmark evidence
The full public result ledger is available for audit without forcing a wide desktop table onto a phone.
Browse raw public benchmark evidence53 rows
Agentic
Terminal-Bench 2.0
DeepSeek V4 Pro 081367.9%SourceKimi K2.666.7%SourceDeepSeek V4 Pro 0813 leads this result
Terminal-Bench 2.1
DeepSeek V4 Pro 081387.9%SourceKimi K2.6—Not directly comparable
BrowseComp
DeepSeek V4 Pro 081383.4%SourceKimi K2.683.2%SourceDeepSeek V4 Pro 0813 leads this result
HLE w/ tools
DeepSeek V4 Pro 081360.0%SourceKimi K2.6—Not directly comparable
MCP Atlas
DeepSeek V4 Pro 081373.6%SourceKimi K2.655.9%SourceDeepSeek V4 Pro 0813 leads this result
Toolathlon
DeepSeek V4 Pro 081351.8%SourceKimi K2.650%SourceDeepSeek V4 Pro 0813 leads this result
CyberGym
DeepSeek V4 Pro 081383.3%SourceKimi K2.6—Not directly comparable
Toolathlon-Verified
DeepSeek V4 Pro 081374.1%SourceKimi K2.6—Not directly comparable
Agents' Last Exam
DeepSeek V4 Pro 081325.7%SourceKimi K2.6—Not directly comparable
AutomationBench
DeepSeek V4 Pro 081331.8%SourceKimi K2.6—Not directly comparable
OSWorld-Verified
DeepSeek V4 Pro 0813—Kimi K2.673.1%SourceNot directly comparable
Claw-Eval
DeepSeek V4 Pro 0813—Kimi K2.662.3%SourceNot directly comparable
DeepSearchQA
DeepSeek V4 Pro 0813—Kimi K2.692.5%SourceNot directly comparable
WideResearch
DeepSeek V4 Pro 0813—Kimi K2.680.8%SourceNot directly comparable
Gert Labs
DeepSeek V4 Pro 0813—Kimi K2.656.82%SourceNot directly comparable
ResearchClawBench
DeepSeek V4 Pro 0813—Kimi K2.618.0%SourceNot directly comparable
OSWorld 2.0
DeepSeek V4 Pro 0813—Kimi K2.64.6%SourceNot directly comparable
Coding
LiveCodeBench Pass@1-COT
DeepSeek V4 Pro 081393.5%SourceKimi K2.6—Not directly comparable
Codeforces
DeepSeek V4 Pro 08133206.0SourceKimi K2.6—Not directly comparable
SWE-bench Verified
DeepSeek V4 Pro 081380.6%SourceKimi K2.680.2%SourceDeepSeek V4 Pro 0813 leads this result
SWE-bench Pro
DeepSeek V4 Pro 081355.4%SourceKimi K2.658.6%SourceKimi K2.6 leads this result
SWE Multilingual
DeepSeek V4 Pro 081376.2%SourceKimi K2.676.7%SourceKimi K2.6 leads this result
Terminal-Bench 2.0
DeepSeek V4 Pro 081367.9%SourceKimi K2.666.7%SourceDeepSeek V4 Pro 0813 leads this result
Vibe Code Bench
Shared sourceDeepSeek V4 Pro 081349.93%Kimi K2.637.89%DeepSeek V4 Pro 0813 leads this result
Terminal-Bench 2.1
DeepSeek V4 Pro 081387.9%SourceKimi K2.6—Not directly comparable
NL2Repo
DeepSeek V4 Pro 081361.5%SourceKimi K2.6—Not directly comparable
deepSwe
DeepSeek V4 Pro 081362.7%SourceKimi K2.6—Not directly comparable
DSBench-FullStack
DeepSeek V4 Pro 081371.1%SourceKimi K2.6—Not directly comparable
DSBench-Hard
DeepSeek V4 Pro 081367.2%SourceKimi K2.6—Not directly comparable
LiveCodeBench v6
DeepSeek V4 Pro 0813—Kimi K2.689.6%SourceNot directly comparable
SciCode
DeepSeek V4 Pro 0813—Kimi K2.652.2%SourceNot directly comparable
cursorBench31
DeepSeek V4 Pro 0813—Kimi K2.647.6%SourceNot directly comparable
Reasoning
Knowledge
MMLU-Pro
DeepSeek V4 Pro 081387.5%SourceKimi K2.6—Not directly comparable
SimpleQA
DeepSeek V4 Pro 081357.9%SourceKimi K2.6—Not directly comparable
Chinese-SimpleQA
DeepSeek V4 Pro 081384.4%SourceKimi K2.6—Not directly comparable
GPQA
DeepSeek V4 Pro 081390.1%SourceKimi K2.690.5%SourceKimi K2.6 leads this result
GPQA-D
DeepSeek V4 Pro 081390.1%SourceKimi K2.690.5%SourceKimi K2.6 leads this result
HLE
DeepSeek V4 Pro 081342.7%SourceKimi K2.634.7%SourceDeepSeek V4 Pro 0813 leads this result
Math
HMMT Feb 2026
DeepSeek V4 Pro 081395.2%SourceKimi K2.692.7%SourceDeepSeek V4 Pro 0813 leads this result
IMOAnswerBench
DeepSeek V4 Pro 081389.8%SourceKimi K2.6—Not directly comparable
Apex
DeepSeek V4 Pro 081338.3%SourceKimi K2.6—Not directly comparable
Apex Shortlist
DeepSeek V4 Pro 081390.2%SourceKimi K2.6—Not directly comparable
AIME26
DeepSeek V4 Pro 0813—Kimi K2.696.4%SourceNot directly comparable
MMAnswerBench
DeepSeek V4 Pro 0813—Kimi K2.686.0%SourceNot directly comparable
FrontierMath v2 (Tiers 1-3)
DeepSeek V4 Pro 0813—Kimi K2.638.966%SourceNot directly comparable
FrontierMath v2 (Tier 4)
DeepSeek V4 Pro 0813—Kimi K2.614.580%SourceNot directly comparable
Multimodal
MMMU-Pro
DeepSeek V4 Pro 0813—Kimi K2.679.4%SourceNot directly comparable
MMMU-Pro w/ Python
DeepSeek V4 Pro 0813—Kimi K2.680.1%SourceNot directly comparable
CharXiv
DeepSeek V4 Pro 0813—Kimi K2.680.4%SourceNot directly comparable
MathVision
DeepSeek V4 Pro 0813—Kimi K2.687.4%SourceNot directly comparable
V*
DeepSeek V4 Pro 0813—Kimi K2.696.9%SourceNot directly comparable
Frequently asked questions
Which is better, DeepSeek V4 Pro 0813 or Kimi K2.6?
DeepSeek V4 Pro 0813 has the higher public score estimate, 60.93 versus 60, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, DeepSeek V4 Pro 0813 or Kimi K2.6?
The current coding averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, DeepSeek V4 Pro 0813 or Kimi K2.6?
The current agentic tasks averages use different weighted benchmark sets, so BenchLM does not name a winner from them. Read the shared benchmark rows directly and test the models on the same task set.
Which costs less, DeepSeek V4 Pro 0813 or Kimi K2.6?
For the stated presets, chat costs $0.00087 on DeepSeek V4 Pro 0813 and $0.00295 on Kimi K2.6; repository review costs $0.02436 and $0.0595; the cache-heavy agent loop costs $0.01812 and $0.249. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, DeepSeek V4 Pro 0813 or Kimi K2.6?
DeepSeek V4 Pro 0813 has the larger documented context window: 1M, compared with 256K.
Related comparisons
Last updated August 13, 2026
Watch DeepSeek V4 Pro 0813 vs Kimi K2.6
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.