Agentic
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- 89.5
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Claude Mythos Preview vs Kimi K3
Updated August 13, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. Use the documented cost, context, and runtime rows instead.
0 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.
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 Mythos Preview 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
No shared weighted benchmark basis supports a winner.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Not enough matched evidence
A complete context comparison is not sourced.
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.
2 Claude Mythos Preview only
40 Kimi K3 only
- Shared results
- 0
- Claude Mythos Preview only
- 2
- Kimi K3 only
- 40
- Like-for-like categories
- 0 / 8
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
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Reasoning
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Knowledge
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- 61.0
- Weighted basis
- 0 vs 2 rows
- Reading
- Not comparable
Math
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multilingual
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
Multimodal
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- 78.5
- Weighted basis
- 0 vs 3 rows
- Reading
- Not comparable
Instruction following
Not comparable- Claude Mythos Preview
- Not measured
- Kimi K3
- Not measured
- Weighted basis
- 0 vs 0 rows
- Reading
- Not comparable
| Category | Claude Mythos Preview | Kimi K3 | Weighted basis | Reading |
|---|---|---|---|---|
| Agentic | Not measured | 89.5 | Not comparable0 vs 2 rows | Not comparable |
| Coding | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Reasoning | Not measured | Not measured | Not comparable0 vs 0 rows | Not comparable |
| Knowledge | Not measured | 61.0 | Not comparable0 vs 2 rows | Not comparable |
| Math | 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 | 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.
A shared-evidence shape is not available.
BenchLM does not draw a radar or infer missing axes when the matched evidence is too sparse.
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 Mythos Preview
- $0.0875
- Fit state unavailable
- 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 Mythos Preview
- $1.63
- Fit state unavailable
- 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 Mythos Preview
- $6.75
- Fit state unavailable
- 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 Mythos Preview 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 Mythos PreviewKimi K3
Context window
Maximum documented context; output-token limits may be lower.
Claude Mythos Preview
Not sourced
Kimi K3
1.05M
API model ID
Claude Mythos Preview
claude-mythos-preview
Anthropic Project GlasswingKimi 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 Mythos Preview
Not published
Anthropic Project GlasswingKimi K3
$0.3 per 1M cached input tokens
Documented inputs
Claude Mythos Preview
Not sourced
Kimi K3
Not sourced
Documented outputs
Claude Mythos Preview
Not sourced
Kimi K3
Not sourced
Provider availability
Claude Mythos Preview
Not sourced
Kimi K3
Not sourced
Reasoning profile
Claude Mythos Preview
Reasoning
Kimi K3
Reasoning
Weight access
Claude Mythos Preview
Proprietary
Kimi K3
Pending
License
Claude Mythos Preview
Proprietary
Kimi K3
Pending
Release date
Claude Mythos Preview
2026-04-07
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
- The public evidence has no benchmark result shared by both models, so it does not support a quality verdict.
- Workload cost
- Repository review: $1.63 vs $0.195. Cache-heavy agent loop: $6.75 vs $0.27.
- Context tradeoff
- A complete documented context comparison is not available.
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 evidence42 rows
Agentic
CyberGym
Claude Mythos Preview83.1%SourceKimi K3—Not directly comparable
ExploitGym
Claude Mythos Preview17.5%SourceKimi K3—Not directly comparable
Terminal-Bench 2.0
Claude Mythos Preview—Kimi K388.3%SourceNot directly comparable
BrowseComp
Claude Mythos Preview—Kimi K391.2%SourceNot directly comparable
DeepSearchQA
Claude Mythos Preview—Kimi K395.0%SourceNot directly comparable
Toolathlon-Verified
Claude Mythos Preview—Kimi K373.2%SourceNot directly comparable
MCP Atlas
Claude Mythos Preview—Kimi K384.2%SourceNot directly comparable
AutomationBench
Claude Mythos Preview—Kimi K330.8%SourceNot directly comparable
JobBench
Claude Mythos Preview—Kimi K352.9%SourceNot directly comparable
APEX-Agents
Claude Mythos Preview—Kimi K337.6%SourceNot directly comparable
SpreadsheetBench 2
Claude Mythos Preview—Kimi K334.8%SourceNot directly comparable
DECK-Bench
Claude Mythos Preview—Kimi K373.5%SourceNot directly comparable
Coding
deepSwe
Claude Mythos Preview—Kimi K367.5%SourceNot directly comparable
cursorBench32
Claude Mythos Preview—Kimi K360.8%SourceNot directly comparable
FrontierSWE
Claude Mythos Preview—Kimi K381.2%SourceNot directly comparable
ProgramBench
Claude Mythos Preview—Kimi K377.8%SourceNot directly comparable
Kimi Code Bench v2
Claude Mythos Preview—Kimi K372.9%SourceNot directly comparable
sweMarathon
Claude Mythos Preview—Kimi K342%SourceNot directly comparable
PostTrain Bench
Claude Mythos Preview—Kimi K336.6%SourceNot directly comparable
MLS-Bench Lite
Claude Mythos Preview—Kimi K348.3%SourceNot directly comparable
VulcanBench v3
Claude Mythos Preview—Kimi K373.9%SourceNot directly comparable
APEX-SWE
Claude Mythos Preview—Kimi K348.0%SourceNot directly comparable
EEBench
Claude Mythos Preview—Kimi K338.3%SourceNot directly comparable
InferenceEval
Claude Mythos Preview—Kimi K339.8%SourceNot directly comparable
KernelBench Internal
Claude Mythos Preview—Kimi K368.5%SourceNot directly comparable
Knowledge
Multimodal
OfficeQA Pro
Claude Mythos Preview—Kimi K363.3%SourceNot directly comparable
MMMU-Pro
Claude Mythos Preview—Kimi K381.6%SourceNot directly comparable
MMMU-Pro w/ Python
Claude Mythos Preview—Kimi K383.4%SourceNot directly comparable
CharXiv w/o tools
Claude Mythos Preview—Kimi K384.8%SourceNot directly comparable
CharXiv
Claude Mythos Preview—Kimi K391.3%SourceNot directly comparable
MathVision
Claude Mythos Preview—Kimi K394.3%SourceNot directly comparable
MathVision w/ Python
Claude Mythos Preview—Kimi K397.8%SourceNot directly comparable
BabyVision w/ Python
Claude Mythos Preview—Kimi K385.7%SourceNot directly comparable
ZeroBench
Claude Mythos Preview—Kimi K323.0%SourceNot directly comparable
ZeroBench w/ Python
Claude Mythos Preview—Kimi K341.0%SourceNot directly comparable
WorldVQA ForceAnswer
Claude Mythos Preview—Kimi K351.0%SourceNot directly comparable
OmniDocBench
Claude Mythos Preview—Kimi K391.1%SourceNot directly comparable
PerceptionBench
Claude Mythos Preview—Kimi K358.5%SourceNot directly comparable
Frequently asked questions
Which is better, Claude Mythos Preview or Kimi K3?
The public evidence has no benchmark result shared by both models, so it does not support a quality verdict. The page therefore keeps the decision tied to the specific documented workload.
Which is better for coding, Claude Mythos Preview 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 Mythos Preview or Kimi K3?
The published evidence does not provide a shared weighted agentic tasks basis for both models, so BenchLM does not name a agentic tasks winner.
Which costs less, Claude Mythos Preview or Kimi K3?
For the stated presets, chat costs $0.0875 on Claude Mythos Preview and $0.0105 on Kimi K3; repository review costs $1.63 and $0.195; the cache-heavy agent loop costs $6.75 and $0.27. Claude Mythos Preview has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Claude Mythos Preview or Kimi K3?
A complete documented context-window comparison is not available.
Related comparisons
Last updated August 13, 2026
Watch Claude Mythos Preview 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.