Knowledge
Like-for-like- Claude Opus 4.8
- 62.7
- Kimi K3
- 61.0
- Weighted basis
- 2 vs 2 rows
- Reading
- Claude Opus 4.8 leads
Claude Opus 4.8 vs Kimi K3
Updated August 19, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Kimi K3 has the higher public score estimate, 80.53 versus 76.61, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
16 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
Kimi K3
Kimi K3 has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Kimi K3
Kimi K3 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
Kimi K3
Kimi K3 has the lower estimated token cost for this stated workload. Claude Opus 4.8 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
Kimi K3
Kimi K3 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
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.
24 Claude Opus 4.8 only
16 Shared
24 Kimi K3 only
- Shared results
- 16
- Claude Opus 4.8 only
- 24
- Kimi K3 only
- 24
- Like-for-like categories
- 1 / 8
2 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.
Agentic
Directional only- Claude Opus 4.8
- 80.3
- Kimi K3
- 89.5
- Weighted basis
- 3 vs 2 rows
- Reading
- Directional only
Multimodal
Directional only- Claude Opus 4.8
- 77.0
- Kimi K3
- 78.5
- Weighted basis
- 2 vs 3 rows
- Reading
- Directional only
Coding
Not comparable- Claude Opus 4.8
- 81.1
- Kimi K3
- Not measured
- Weighted basis
- 2 vs 0 rows
- Reading
- Not comparable
Reasoning
Not comparable- Claude Opus 4.8
- 72.1
- Kimi K3
- Not measured
- Weighted basis
- 1 vs 0 rows
- Reading
- Not comparable
Math
Not comparable- Claude Opus 4.8
- 53.9
- Kimi K3
- Not measured
- Weighted basis
- 3 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Opus 4.8
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Opus 4.8
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Claude Opus 4.8 | Kimi K3 | Weighted basis | Reading |
|---|---|---|---|---|
| Knowledge | 62.7 | 61.0 | Like-for-like2 vs 2 rows | Claude Opus 4.8 leads |
| Agentic | 80.3 | 89.5 | Directional only3 vs 2 rows | Directional only |
| Multimodal | 77.0 | 78.5 | Directional only2 vs 3 rows | Directional only |
| Coding | 81.1 | Not measured | Not comparable2 vs 0 rows | Not comparable |
| Reasoning | 72.1 | Not measured | Not comparable1 vs 0 rows | Not comparable |
| Math | 53.9 | Not measured | Not comparable3 vs 0 rows | Not comparable |
| Multilingual | Not measured | Not measured | Not comparable0 vs 0 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
BrowseComp
Agentic
OfficeQA Pro
Multimodal
HLE
Knowledge
CharXiv
Multimodal
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
- Claude Opus 4.8
- $0.0175
- Fits in one request
- Kimi K3
- $0.0105
- Fits in one request
Kimi K3 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Claude Opus 4.8
- $0.325
- Fits in one request
- Kimi K3
- $0.195
- Fits in one request
Kimi K3 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Claude Opus 4.8
- $1.35
- Fits in one request
- Cached input priced at the published list-input rate
- Kimi K3
- $0.27
- Fits in one request
Kimi K3 has the lower modeled cost
Claude Opus 4.8 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.
SpecificationClaude Opus 4.8Kimi K3
Context window
Maximum documented context; output-token limits may be lower.
Claude Opus 4.8
Kimi K3
1.05M
API model ID
Claude Opus 4.8
claude-opus-4-8
Anthropic model overviewKimi K3
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Claude Opus 4.8
Not published
Kimi K3
$0.3 per 1M cached input tokens
Documented inputs
Claude Opus 4.8
text, image
Anthropic model overviewKimi K3
Not sourced
Documented outputs
Claude Opus 4.8
Kimi K3
Not sourced
Provider availability
Claude Opus 4.8
Generally Available · Claude API
Anthropic model overviewKimi K3
Not sourced
Reasoning profile
Claude Opus 4.8
Reasoning
Kimi K3
Reasoning
Weight access
Claude Opus 4.8
Proprietary
Kimi K3
Pending
License
Claude Opus 4.8
Proprietary
Kimi K3
Pending
Release date
Claude Opus 4.8
2026-05-28
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
- Kimi K3 has the higher public score estimate, 80.53 versus 76.61, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.325 vs $0.195. Cache-heavy agent loop: $1.35 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 3.0
Claude Opus 4.821.1%SourceKimi K3—Not directly comparable
Terminal-Bench 2.0
Claude Opus 4.874.6%SourceKimi K388.3%SourceKimi K3 leads this result
BrowseComp
Claude Opus 4.884.3%SourceKimi K391.2%SourceKimi K3 leads this result
DeepSearchQA
Claude Opus 4.893.1%SourceKimi K395.0%SourceKimi K3 leads this result
OSWorld-Verified
Claude Opus 4.883.4%SourceKimi K3—Not directly comparable
Finance Agent v2
Claude Opus 4.853.9%SourceKimi K3—Not directly comparable
MCP Atlas
Claude Opus 4.882.2%SourceKimi K384.2%SourceKimi K3 leads this result
Toolathlon
Claude Opus 4.859.9%SourceKimi K3—Not directly comparable
Gert Labs
Claude Opus 4.872.97%SourceKimi K3—Not directly comparable
ResearchClawBench
Claude Opus 4.821.1%SourceKimi K3—Not directly comparable
OSWorld 2.0
Claude Opus 4.820.6%SourceKimi K3—Not directly comparable
Toolathlon-Verified
Claude Opus 4.8—Kimi K373.2%SourceNot directly comparable
AutomationBench
Claude Opus 4.8—Kimi K330.8%SourceNot directly comparable
JobBench
Claude Opus 4.8—Kimi K352.9%SourceNot directly comparable
APEX-Agents
Claude Opus 4.8—Kimi K337.6%SourceNot directly comparable
SpreadsheetBench 2
Claude Opus 4.8—Kimi K334.8%SourceNot directly comparable
DECK-Bench
Claude Opus 4.8—Kimi K373.5%SourceNot directly comparable
Coding
SWE-bench Verified
Claude Opus 4.888.6%SourceKimi K3—Not directly comparable
SWE-bench Pro
Claude Opus 4.869.2%SourceKimi K3—Not directly comparable
SWE Multilingual
Claude Opus 4.884.4%SourceKimi K3—Not directly comparable
SWE Multimodal
Claude Opus 4.838.4%SourceKimi K3—Not directly comparable
Terminal-Bench 2.0
Claude Opus 4.874.6%SourceKimi K3—Not directly comparable
cursorBench31
Claude Opus 4.858.4%SourceKimi K3—Not directly comparable
cursorBench32
Shared sourceClaude Opus 4.862.3%Kimi K360.8%Claude Opus 4.8 leads this result
FrontierCode 1.1 Main
Claude Opus 4.846.5%SourceKimi K3—Not directly comparable
APEX-SWE
Shared sourceClaude Opus 4.843.9%Kimi K348.0%Kimi K3 leads this result
EEBench
Shared sourceClaude Opus 4.851.4%Kimi K338.3%Claude Opus 4.8 leads this result
3DCodeBench
Claude Opus 4.847.0%SourceKimi K3—Not directly comparable
CADGenBench Generation
Claude Opus 4.827.4%SourceKimi K3—Not directly comparable
SpaceXAI MTS Eval
Claude Opus 4.855.8%SourceKimi K3—Not directly comparable
InferenceEval
Shared sourceClaude Opus 4.840.7%Kimi K339.8%Claude Opus 4.8 leads this result
KernelBench Internal
Shared sourceClaude Opus 4.867.6%Kimi K368.5%Kimi K3 leads this result
deepSwe
Claude Opus 4.8—Kimi K367.5%SourceNot directly comparable
FrontierSWE
Claude Opus 4.8—Kimi K381.2%SourceNot directly comparable
ProgramBench
Claude Opus 4.8—Kimi K377.8%SourceNot directly comparable
Kimi Code Bench v2
Claude Opus 4.8—Kimi K372.9%SourceNot directly comparable
sweMarathon
Claude Opus 4.8—Kimi K342%SourceNot directly comparable
PostTrain Bench
Claude Opus 4.8—Kimi K336.6%SourceNot directly comparable
MLS-Bench Lite
Claude Opus 4.8—Kimi K348.3%SourceNot directly comparable
VulcanBench v3
Claude Opus 4.8—Kimi K373.9%SourceNot directly comparable
Reasoning
Knowledge
GPQA
Claude Opus 4.893.6%SourceKimi K393.5%SourceClaude Opus 4.8 leads this result
GPQA-D
Claude Opus 4.893.6%SourceKimi K393.5%SourceClaude Opus 4.8 leads this result
HLE
Claude Opus 4.857.9%SourceKimi K356%SourceClaude Opus 4.8 leads this result
HLE w/o tools
Claude Opus 4.849.8%SourceKimi K343.5%SourceClaude Opus 4.8 leads this result
Math
Multilingual
INCLUDE
Claude Opus 4.887.6%SourceKimi K3—Not directly comparable
Multimodal
OfficeQA Pro
Claude Opus 4.866.2%SourceKimi K363.3%SourceClaude Opus 4.8 leads this result
ScreenSpot Pro
Claude Opus 4.887.9%SourceKimi K3—Not directly comparable
CharXiv
Claude Opus 4.889.9%SourceKimi K391.3%SourceKimi K3 leads this result
CharXiv w/o tools
Claude Opus 4.880.5%SourceKimi K384.8%SourceKimi K3 leads this result
MMMU-Pro
Claude Opus 4.8—Kimi K381.6%SourceNot directly comparable
MMMU-Pro w/ Python
Claude Opus 4.8—Kimi K383.4%SourceNot directly comparable
MathVision
Claude Opus 4.8—Kimi K394.3%SourceNot directly comparable
MathVision w/ Python
Claude Opus 4.8—Kimi K397.8%SourceNot directly comparable
BabyVision w/ Python
Claude Opus 4.8—Kimi K385.7%SourceNot directly comparable
ZeroBench
Claude Opus 4.8—Kimi K323.0%SourceNot directly comparable
ZeroBench w/ Python
Claude Opus 4.8—Kimi K341.0%SourceNot directly comparable
WorldVQA ForceAnswer
Claude Opus 4.8—Kimi K351.0%SourceNot directly comparable
OmniDocBench
Claude Opus 4.8—Kimi K391.1%SourceNot directly comparable
PerceptionBench
Claude Opus 4.8—Kimi K358.5%SourceNot directly comparable
Frequently asked questions
Which is better, Claude Opus 4.8 or Kimi K3?
Kimi K3 has the higher public score estimate, 80.53 versus 76.61, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Claude Opus 4.8 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, Claude Opus 4.8 or Kimi K3?
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, Claude Opus 4.8 or Kimi K3?
For the stated presets, chat costs $0.0175 on Claude Opus 4.8 and $0.0105 on Kimi K3; repository review costs $0.325 and $0.195; the cache-heavy agent loop costs $1.35 and $0.27. Claude Opus 4.8 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Claude Opus 4.8 or Kimi K3?
Kimi K3 has the larger documented context window: 1.05M, compared with 1M.
Related comparisons
Last updated August 19, 2026
Watch Claude Opus 4.8 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.