Multimodal
Directional only- Kimi K2.6
- 63.9
- #26/49
- Step 5 Preview
- 61.2
- #28/49
- Basis
- Provisional lane · 2 vs 2 weighted rows
- Reading
- Directional only
Kimi K2.6 vs Step 5 Preview
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.
6 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
Updated September 21, 2026. We do not rank this pair: at least one has no public score. Public scores include evidence status and uncertainty.
Both of these models will change. Get the price, version and retirement notices for the pair, each with its source. Follow model changes
Share or export
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
Step 5 Preview
Step 5 Preview has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Step 5 Preview
Step 5 Preview 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
Step 5 Preview
Step 5 Preview 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
Step 5 Preview
Step 5 Preview 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
Step 5 Preview is not ranked on the public lane for coding, so no winner is named for coding.
Confidence: limited
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Step 5 Preview is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limited
Which one for a specific job
Choose a job from the LLM Selector's task catalog to see the category row it rests on, under the same basis rules as the table below. A directional row stays directional; choosing a job never creates a winner.
The same task catalog as the LLM Selector. Each job names the evidence surface it rests on; nothing here adds a new score.
51.5Kimi K2.6—Step 5 Preview
Not comparable · BenchAlign
The coding row is not comparable on the public lane: at least one model is not measured or not ranked there.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
31 Kimi K2.6 only
6 Shared
16 Step 5 Preview only
- Shared results
- 6
- Kimi K2.6 only
- 31
- Step 5 Preview only
- 16
- Like-for-like categories
- 0 / 8
1 category rests on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign lane for agentic, coding, and knowledge, and the weighted public rows elsewhere, with the same rank each model page reports. A row is like-for-like only when both scores rest on Supported evidence or the same weighted set. Directional and not-comparable rows remain visible, but they never receive a winner.
Agentic
Not comparable- Kimi K2.6
- 47.6
- Supported · #73/154
- Step 5 Preview
- Not ranked
- Basis
- BenchAlign lane · 12 vs 11 public rows
- Reading
- Not comparable
Coding
Not comparable- Kimi K2.6
- 51.5
- Supported · #54/156
- Step 5 Preview
- Not ranked
- Basis
- BenchAlign lane · 10 vs 6 public rows
- Reading
- Not comparable
Reasoning
Not comparable- Kimi K2.6
- Not ranked
- Step 5 Preview
- 78.5
- Unranked · 2 rankable rows
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Knowledge
Not comparable- Kimi K2.6
- 61.5
- Supported · #32/186
- Step 5 Preview
- Not ranked
- Basis
- BenchAlign lane · 5 vs 2 public rows
- Reading
- Not comparable
Multilingual
Not comparable- Kimi K2.6
- Not ranked
- Step 5 Preview
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Kimi K2.6
- Not ranked
- Step 5 Preview
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- Kimi K2.6
- 71.0
- #1/7
- Step 5 Preview
- Not ranked
- Basis
- Provisional lane · 4 vs 0 weighted rows
- Reading
- Not comparable
| Category | Kimi K2.6 | Step 5 Preview | Basis | Reading |
|---|---|---|---|---|
| Multimodal | 63.9#26/49 | 61.2#28/49 | Directional onlyProvisional lane · 2 vs 2 weighted rows | Directional only |
| Agentic | 47.6Supported · #73/154 | Not ranked | Not comparableBenchAlign lane · 12 vs 11 public rows | Not comparable |
| Coding | 51.5Supported · #54/156 | Not ranked | Not comparableBenchAlign lane · 10 vs 6 public rows | Not comparable |
| Reasoning | Not ranked | 78.5Unranked · 2 rankable rows | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Knowledge | 61.5Supported · #32/186 | Not ranked | Not comparableBenchAlign lane · 5 vs 2 public rows | Not comparable |
| Multilingual | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Instruction following | Not ranked | Not ranked | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | 71.0#1/7 | Not ranked | Not comparableProvisional lane · 4 vs 0 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign) differ from the provisional-lane categories. Unranked scores sit on the lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
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
SciCode
Coding
BrowseComp
Agentic
MMMU-Pro
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
- Kimi K2.6
- $0.00295
- Fits in one request
- Step 5 Preview
- $0.00235
- Fits in one request
Step 5 Preview has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Kimi K2.6
- $0.0595
- Fits in one request
- Step 5 Preview
- $0.0581
- Fits in one request
Step 5 Preview has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Kimi K2.6
- $0.249
- Fits in one request
- Cached input priced at the published list-input rate
- Step 5 Preview
- $0.057
- Fits in one request
Step 5 Preview 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.
SpecificationKimi K2.6Step 5 Preview
Context window
Maximum documented context; output-token limits may be lower.
Kimi K2.6
256K
Step 5 Preview
API model ID
Kimi K2.6
Not sourced
Step 5 Preview
step-5-preview
StepFun Step 5 Preview model documentationCached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Kimi K2.6
Not published
Step 5 Preview
$0.05 per 1M cached input tokens
StepFun pricing and rate limitsDocumented inputs
Kimi K2.6
Not sourced
Step 5 Preview
text, image, video
StepFun Step 5 Preview model documentationDocumented outputs
Kimi K2.6
Not sourced
Step 5 Preview
Provider availability
Kimi K2.6
Not sourced
Step 5 Preview
Preview · StepFun Open Platform API
StepFun Step 5 Preview model documentationReasoning profile
Kimi K2.6
Reasoning
Step 5 Preview
Reasoning
Weight access
Kimi K2.6
Open Weight
Step 5 Preview
Pending
License
Kimi K2.6
Open Weight
Step 5 Preview
Pending
Release date
Kimi K2.6
2026-04-20
Step 5 Preview
2026-09-20
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
- 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: $0.0595 vs $0.0581. Cache-heavy agent loop: $0.249 vs $0.057.
- Context tradeoff
- Step 5 Preview 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.
Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/day
Step 5 Preview
API / mo$2,775
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
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
Kimi K2.666.7%SourceStep 5 Preview—Not directly comparable
BrowseComp
Kimi K2.683.2%SourceStep 5 Preview88.7%SourceStep 5 Preview leads this result
OSWorld-Verified
Kimi K2.673.1%SourceStep 5 Preview—Not directly comparable
Toolathlon
Kimi K2.650%SourceStep 5 Preview—Not directly comparable
MCP Atlas
Kimi K2.655.9%SourceStep 5 Preview85.6%SourceStep 5 Preview leads this result
Claw-Eval
Kimi K2.662.3%SourceStep 5 Preview—Not directly comparable
DeepSearchQA
Kimi K2.692.5%SourceStep 5 Preview—Not directly comparable
WideResearch
Kimi K2.680.8%SourceStep 5 Preview—Not directly comparable
Gert Labs
Kimi K2.656.82%SourceStep 5 Preview—Not directly comparable
ResearchClawBench
Kimi K2.618.0%SourceStep 5 Preview—Not directly comparable
OSWorld 2.0
Kimi K2.64.6%SourceStep 5 Preview—Not directly comparable
Terminal-Bench 2.1 (Vals)
Kimi K2.653.6%SourceStep 5 Preview—Not directly comparable
Terminal-Bench 2.1
Kimi K2.6—Step 5 Preview85.0%SourceNot directly comparable
Terminal-Bench 4.0
Kimi K2.6—Step 5 Preview33.30%SourceNot directly comparable
CyberGym
Kimi K2.6—Step 5 Preview84.7%SourceNot directly comparable
AutomationBench
Kimi K2.6—Step 5 Preview44.0%SourceNot directly comparable
Toolathlon-Verified
Kimi K2.6—Step 5 Preview74.1%SourceNot directly comparable
JobBench
Kimi K2.6—Step 5 Preview59.0%SourceNot directly comparable
APEX-Agents
Kimi K2.6—Step 5 Preview37.8%SourceNot directly comparable
DRACO
Kimi K2.6—Step 5 Preview83.3%SourceNot directly comparable
Agents' Last Exam
Kimi K2.6—Step 5 Preview29.5%SourceNot directly comparable
Coding
SWE-bench Verified
Kimi K2.680.2%SourceStep 5 Preview—Not directly comparable
LiveCodeBench v6
Kimi K2.689.6%SourceStep 5 Preview—Not directly comparable
SWE-bench Pro
Kimi K2.658.6%SourceStep 5 Preview—Not directly comparable
SWE Multilingual
Kimi K2.676.7%SourceStep 5 Preview—Not directly comparable
SciCode
Kimi K2.652.2%SourceStep 5 Preview58.9%SourceStep 5 Preview leads this result
Terminal-Bench 2.0
Kimi K2.666.7%SourceStep 5 Preview—Not directly comparable
Vibe Code Bench
Kimi K2.637.89%SourceStep 5 Preview—Not directly comparable
cursorBench31
Kimi K2.647.6%SourceStep 5 Preview—Not directly comparable
LiveCodeBench (Vals)
Kimi K2.686.8%SourceStep 5 Preview—Not directly comparable
SWE-bench (Vals)
Kimi K2.676.2%SourceStep 5 Preview—Not directly comparable
DeepSWE
Kimi K2.6—Step 5 Preview67.7%SourceNot directly comparable
Terminal-Bench 2.1
Kimi K2.6—Step 5 Preview85.0%SourceNot directly comparable
ProgramBench
Kimi K2.6—Step 5 Preview80.5%SourceNot directly comparable
sweMarathon
Kimi K2.6—Step 5 Preview72.7%SourceNot directly comparable
MLS-Bench Lite
Kimi K2.6—Step 5 Preview40.5%SourceNot directly comparable
Reasoning
CritPt
Kimi K2.6—Step 5 Preview20.9%SourceNot directly comparable
Multimodal
MMMU-Pro
Kimi K2.679.4%SourceStep 5 Preview76%SourceKimi K2.6 leads this result
MMMU-Pro w/ Python
Kimi K2.680.1%SourceStep 5 Preview—Not directly comparable
CharXiv
Kimi K2.680.4%SourceStep 5 Preview—Not directly comparable
MathVision
Kimi K2.687.4%SourceStep 5 Preview—Not directly comparable
V*
Kimi K2.696.9%SourceStep 5 Preview—Not directly comparable
OfficeQA Pro
Kimi K2.6—Step 5 Preview60.3%SourceNot directly comparable
Knowledge
GPQA
Kimi K2.690.5%SourceStep 5 Preview—Not directly comparable
GPQA-D
Kimi K2.690.5%SourceStep 5 Preview93.5%SourceStep 5 Preview leads this result
HLE
Kimi K2.634.7%SourceStep 5 Preview46.5%SourceStep 5 Preview leads this result
GPQA Diamond (Vals)
Kimi K2.689.1%SourceStep 5 Preview—Not directly comparable
MMLU-Pro (Vals)
Kimi K2.687.6%SourceStep 5 Preview—Not directly comparable
Math
AIME26
Kimi K2.696.4%SourceStep 5 Preview—Not directly comparable
HMMT Feb 2026
Kimi K2.692.7%SourceStep 5 Preview—Not directly comparable
MMAnswerBench
Kimi K2.686.0%SourceStep 5 Preview—Not directly comparable
FrontierMath v2 (Tiers 1-3)
Kimi K2.638.966%SourceStep 5 Preview—Not directly comparable
FrontierMath v2 (Tier 4)
Kimi K2.614.580%SourceStep 5 Preview—Not directly comparable
Questions
Which is better, Kimi K2.6 or Step 5 Preview?
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, Kimi K2.6 or Step 5 Preview?
Step 5 Preview is not ranked on the public lane for coding, so no winner is named for coding.
Which is better for agentic tasks, Kimi K2.6 or Step 5 Preview?
Step 5 Preview is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.
Which costs less, Kimi K2.6 or Step 5 Preview?
For the stated presets, chat costs $0.00295 on Kimi K2.6 and $0.00235 on Step 5 Preview; repository review costs $0.0595 and $0.0581; the cache-heavy agent loop costs $0.249 and $0.057. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Kimi K2.6 or Step 5 Preview?
Step 5 Preview has the larger documented context window: 1M, compared with 256K.
Related comparisons
Last updated September 21, 2026
Watch Kimi K2.6 vs Step 5 Preview
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.