Chat turn cost
1K fresh input + 500 output tokens
Kimi K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Gemini 4 Carbon vs Kimi K2.5
Updated October 28, 2026. Rank says Gemini 4 Carbon is ahead. Price, access, and your workload can each overturn that. Public scores include evidence status and uncertainty.
Share or export
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 resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 0
- Gemini 4 Carbon only
- 9
- Kimi K2.5 only
- 68
- Like-for-like categories
- 0 / 8
Estimated: Gemini 4 Carbon · Supported: Kimi K2.5. Conditional ranges do not establish rank confidence.How the comparison works
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.
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Gemini 4 Carbon
Gemini 4 Carbon has the lower estimated token cost for this stated workload. Kimi K2.5 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 K2.5
Kimi K2.5 has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-ratesCoding work
Code generation, repair, and software-engineering tasks
Not enough matched evidence
Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like.
Confidence: limitedAgentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Gemini 4 Carbon is not ranked on the public lane for agentic, so no winner is named for agentic.
Confidence: limitedLong documents
Prompts that approach the documented context limit
Not enough matched evidence
A complete context comparison is not sourced.
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.
82.2Gemini 4 Carbon39.6Kimi K2.5
Directional only · BenchAlign v5.8
Gemini 4 Carbon has the higher coding point estimate. Conditional score ranges do not establish rank confidence.
Coding scores combine specific tasks and setups. Match the editor, harness, and effort to your workflow.
Same basis rules as the category table below
What is actually comparable
Shared results can support a head-to-head reading. Results present for only one model describe coverage, not superiority.
9 Gemini 4 Carbon only
68 Kimi K2.5 only
2 categories rest on Estimated evidence or different evidence sets. Those rows remain visible for coverage context but do not name a winner.
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.
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.8 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.
Coding
Directional only- Gemini 4 Carbon
- 82.2
- Supported · #3/156
- Kimi K2.5
- 39.6
- Estimated · #71/156
- Basis
- BenchAlign v5.8 lane · 1 vs 10 public rows
- Reading
- Directional only
Knowledge
Directional only- Gemini 4 Carbon
- 88.2
- Supported · #2/178
- Kimi K2.5
- 50.4
- Estimated · #78/178
- Basis
- BenchAlign v5.8 lane · 0 vs 8 public rows
- Reading
- Directional only
Agentic
Not comparable- Gemini 4 Carbon
- Not ranked
- Kimi K2.5
- 42.3
- Estimated · #69/134
- Basis
- BenchAlign v5.8 lane · 4 vs 16 public rows
- Reading
- Not comparable
Reasoning
Not comparable- Gemini 4 Carbon
- 79.5
- Unranked · 2 rankable rows
- Kimi K2.5
- 71.3
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Gemini 4 Carbon
- Not ranked
- Kimi K2.5
- 70.6
- #36/58
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Gemini 4 Carbon
- Not ranked
- Kimi K2.5
- 87.3
- #8/17
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Gemini 4 Carbon
- Not ranked
- Kimi K2.5
- 85.8
- #51/157
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- Gemini 4 Carbon
- Not ranked
- Kimi K2.5
- 83.3
- Unranked · 7 rankable rows
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Not comparable
| Category | Gemini 4 Carbon | Kimi K2.5 | Basis | Reading |
|---|---|---|---|---|
| Coding | 82.2Supported · #3/156 | 39.6Estimated · #71/156 | Directional onlyBenchAlign v5.8 lane · 1 vs 10 public rows | Directional only |
| Knowledge | 88.2Supported · #2/178 | 50.4Estimated · #78/178 | Directional onlyBenchAlign v5.8 lane · 0 vs 8 public rows | Directional only |
| Agentic | Not ranked | 42.3Estimated · #69/134 | Not comparableBenchAlign v5.8 lane · 4 vs 16 public rows | Not comparable |
| Reasoning | 79.5Unranked · 2 rankable rows | 71.3Unranked · 3 rankable rows | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Multimodal | Not ranked | 70.6#36/58 | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
| Multilingual | Not ranked | 87.3#8/17 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Instruction following | Not ranked | 85.8#51/157 | Not comparableProvisional lane · 0 vs 0 weighted rows | Not comparable |
| Math | Not ranked | 83.3Unranked · 7 rankable rows | Not comparableProvisional lane · 0 vs 2 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.8) differ from the provisional-lane categories. Unranked scores sit on the provisional lane’s scale but never name a winner. Like-for-like rows within 0.5 points read as a practical tie.
Supported evidence per lane · bars run 0–100Methodology
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
- Gemini 4 Carbon
- $0.007
- Fit state unavailable
- Kimi K2.5
- $0.0021
- Fits in one request
Kimi K2.5 has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Gemini 4 Carbon
- $0.13
- Fit state unavailable
- Kimi K2.5
- $0.039
- Fits in one request
Kimi K2.5 has the lower modeled cost
Costs use the listed standard API rates.
Cache-heavy agent loop
200K cached + 20K fresh input + 10K output tokens
- Gemini 4 Carbon
- $0.16
- Fit state unavailable
- Kimi K2.5
- $0.162
- Fits in one request
- Cached input priced at the published list-input rate
Gemini 4 Carbon has the lower modeled cost
Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
Cached input falls back to the list input rate only where a cached rate is unpublished
Specification differences
Sourced differences are shown directly. Missing facts stay explicit instead of being inferred from a model name or family.
SpecificationGemini 4 CarbonKimi K2.5
Context window
Maximum documented context; output-token limits may be lower.
Gemini 4 Carbon
Not sourced
Google Gemini 4 Carbon launchKimi K2.5
256K
API model ID
Gemini 4 Carbon
Not sourced
Kimi K2.5
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Gemini 4 Carbon
$0.1 per 1M cached input tokens
Google Gemini 4 Carbon launch postKimi K2.5
Not published
Documented inputs
Gemini 4 Carbon
Not sourced
Kimi K2.5
Not sourced
Documented outputs
Gemini 4 Carbon
Not sourced
Kimi K2.5
Not sourced
Provider availability
Gemini 4 Carbon
Limited Access · Fairwind Program
Google Gemini 4 Carbon launchKimi K2.5
Not sourced
Reasoning profile
Gemini 4 Carbon
Reasoning
Kimi K2.5
Non-Reasoning
Weight access
Gemini 4 Carbon
Proprietary
Kimi K2.5
Open Weight
License
Gemini 4 Carbon
Proprietary
Kimi K2.5
Open Weight
Release date
Gemini 4 Carbon
2026-10-28
Kimi K2.5
2026-01-27
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: $0.13 vs $0.039. Cache-heavy agent loop: $0.16 vs $0.162.
- Context tradeoff
- A complete documented context comparison is not available.
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, Gemini 4 Carbon or Kimi K2.5?
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, Gemini 4 Carbon or Kimi K2.5?
Gemini 4 Carbon scores higher for coding on the public lane, 82.2 to 39.6. Kimi K2.5 is scored on Estimated evidence for coding, so the reading is directional rather than like-for-like. BenchLM does not name a winner for coding from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which is better for agentic tasks, Gemini 4 Carbon or Kimi K2.5?
Gemini 4 Carbon is not ranked on the public lane for agentic tasks, so no winner is named for agentic tasks.
Which costs less, Gemini 4 Carbon or Kimi K2.5?
For the stated presets, chat costs $0.007 on Gemini 4 Carbon and $0.0021 on Kimi K2.5; repository review costs $0.13 and $0.039; the cache-heavy agent loop costs $0.16 and $0.162. Kimi K2.5 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Gemini 4 Carbon or Kimi K2.5?
A complete documented context-window comparison is not available.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Gemini 4 Carbon
API / mo$9,000
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.5
API / mo$2,700
Self-host / mo$27,962
Break-even706M/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 evidence77 rows
Agentic
AutomationBench
Gemini 4 Carbon53.9%SourceKimi K2.5—Not directly comparable
OSWorld 2.1 Offline
Gemini 4 Carbon81.5%SourceKimi K2.5—Not directly comparable
CWE-bench v1
Gemini 4 Carbon71.0%SourceKimi K2.5—Not directly comparable
ExploitGym
Gemini 4 Carbon53.5%SourceKimi K2.5—Not directly comparable
Terminal-Bench 2.0
Gemini 4 Carbon—Kimi K2.550.8%SourceNot directly comparable
BrowseComp
Gemini 4 Carbon—Kimi K2.560.6%SourceNot directly comparable
OSWorld-Verified
Gemini 4 Carbon—Kimi K2.563.3%SourceNot directly comparable
Claw-Eval
Gemini 4 Carbon—Kimi K2.552.3%SourceNot directly comparable
QwenClawBench
Gemini 4 Carbon—Kimi K2.554.3%SourceNot directly comparable
τ³-bench results
Gemini 4 Carbon—Kimi K2.565.7%SourceNot directly comparable
DeepSearchQA
Gemini 4 Carbon—Kimi K2.577.1%SourceNot directly comparable
DeepPlanning
Gemini 4 Carbon—Kimi K2.514.4%SourceNot directly comparable
Toolathlon
Gemini 4 Carbon—Kimi K2.527.8%SourceNot directly comparable
MCP Atlas
Gemini 4 Carbon—Kimi K2.529.5%SourceNot directly comparable
MCP-Tasks
Gemini 4 Carbon—Kimi K2.559.1%SourceNot directly comparable
WideResearch
Gemini 4 Carbon—Kimi K2.572.7%SourceNot directly comparable
BFCL v4
Gemini 4 Carbon—Kimi K2.568.3%SourceNot directly comparable
Gert Labs
Gemini 4 Carbon—Kimi K2.545.88%SourceNot directly comparable
ResearchClawBench
Gemini 4 Carbon—Kimi K2.514.0%SourceNot directly comparable
JobBench
Gemini 4 Carbon—Kimi K2.58.7%SourceNot directly comparable
Coding
FrontierSWE v2
Gemini 4 Carbon61.2%SourceKimi K2.5—Not directly comparable
SWE-bench Verified
Gemini 4 Carbon—Kimi K2.576.8%SourceNot directly comparable
SWE-bench Verified*
Gemini 4 Carbon—Kimi K2.570.8%SourceNot directly comparable
LiveCodeBench v6
Gemini 4 Carbon—Kimi K2.585.0%SourceNot directly comparable
SWE-bench Pro
Gemini 4 Carbon—Kimi K2.550.7%SourceNot directly comparable
SWE Multilingual
Gemini 4 Carbon—Kimi K2.573%SourceNot directly comparable
SWE-Rebench
Gemini 4 Carbon—Kimi K2.558.5%SourceNot directly comparable
React Native Evals
Gemini 4 Carbon—Kimi K2.577.2%SourceNot directly comparable
NL2Repo
Gemini 4 Carbon—Kimi K2.532%SourceNot directly comparable
SciCode
Gemini 4 Carbon—Kimi K2.548.7%SourceNot directly comparable
Terminal-Bench 2.0
Gemini 4 Carbon—Kimi K2.550.8%SourceNot directly comparable
Reasoning
Multimodal
Chartography (no tools)
Gemini 4 Carbon72.5%SourceKimi K2.5—Not directly comparable
Blueprint-Bench 2
Gemini 4 Carbon58.6%SourceKimi K2.5—Not directly comparable
MMMU
Gemini 4 Carbon—Kimi K2.584.3%SourceNot directly comparable
MMMU-Pro
Gemini 4 Carbon—Kimi K2.578.5%SourceNot directly comparable
RealWorldQA
Gemini 4 Carbon—Kimi K2.581.0%SourceNot directly comparable
OmniDocBench 1.5
Gemini 4 Carbon—Kimi K2.588.8%SourceNot directly comparable
Video-MME
Gemini 4 Carbon—Kimi K2.587.4%SourceNot directly comparable
Video-MME (with subtitle)
Gemini 4 Carbon—Kimi K2.587.4%SourceNot directly comparable
Video-MME (w/o subtitle)
Gemini 4 Carbon—Kimi K2.583.2%SourceNot directly comparable
MathVision
Gemini 4 Carbon—Kimi K2.584.2%SourceNot directly comparable
We-Math
Gemini 4 Carbon—Kimi K2.584.7%SourceNot directly comparable
DynaMath
Gemini 4 Carbon—Kimi K2.584.4%SourceNot directly comparable
SimpleVQA
Gemini 4 Carbon—Kimi K2.571.2%SourceNot directly comparable
MMLongBench-Doc
Gemini 4 Carbon—Kimi K2.558.5%SourceNot directly comparable
CC-OCR
Gemini 4 Carbon—Kimi K2.579.7%SourceNot directly comparable
AI2D_TEST
Gemini 4 Carbon—Kimi K2.590.8%SourceNot directly comparable
CountBench
Gemini 4 Carbon—Kimi K2.594.1%SourceNot directly comparable
RefCOCO (avg)
Gemini 4 Carbon—Kimi K2.587.8%SourceNot directly comparable
MMVU
Gemini 4 Carbon—Kimi K2.580.4%SourceNot directly comparable
VideoMMMU
Gemini 4 Carbon—Kimi K2.586.6%SourceNot directly comparable
MLVU (M-Avg)
Gemini 4 Carbon—Kimi K2.585.0%SourceNot directly comparable
CharXiv
Gemini 4 Carbon—Kimi K2.577.5%SourceNot directly comparable
Knowledge
GPQA
Gemini 4 Carbon—Kimi K2.587.6%SourceNot directly comparable
GPQA-D
Gemini 4 Carbon—Kimi K2.587.6%SourceNot directly comparable
SuperGPQA
Gemini 4 Carbon—Kimi K2.569.2%SourceNot directly comparable
MMLU-Pro
Gemini 4 Carbon—Kimi K2.587.1%SourceNot directly comparable
MMLU-Pro (Arcee)
Gemini 4 Carbon—Kimi K2.587.1%SourceNot directly comparable
MMLU-Redux
Gemini 4 Carbon—Kimi K2.594.5%SourceNot directly comparable
C-Eval
Gemini 4 Carbon—Kimi K2.594%SourceNot directly comparable
HLE
Gemini 4 Carbon—Kimi K2.530.1%SourceNot directly comparable
Multilingual
MMLU-ProX
Gemini 4 Carbon—Kimi K2.582.3%SourceNot directly comparable
NOVA-63
Gemini 4 Carbon—Kimi K2.556.0%SourceNot directly comparable
INCLUDE
Gemini 4 Carbon—Kimi K2.583.3%SourceNot directly comparable
PolyMath
Gemini 4 Carbon—Kimi K2.543.1%SourceNot directly comparable
MAXIFE
Gemini 4 Carbon—Kimi K2.572.8%SourceNot directly comparable
Instruction following
Math
AIME 2025
Gemini 4 Carbon—Kimi K2.596.1%SourceNot directly comparable
AIME26
Gemini 4 Carbon—Kimi K2.595.8%SourceNot directly comparable
AIME25 (Arcee)
Gemini 4 Carbon—Kimi K2.596.3%SourceNot directly comparable
HMMT Feb 2025
Gemini 4 Carbon—Kimi K2.595.4%SourceNot directly comparable
HMMT Nov 2025
Gemini 4 Carbon—Kimi K2.591.1%SourceNot directly comparable
HMMT Feb 2026
Gemini 4 Carbon—Kimi K2.587.1%SourceNot directly comparable
MMAnswerBench
Gemini 4 Carbon—Kimi K2.581.8%SourceNot directly comparable
77 public results · 0 shared
Watch Gemini 4 Carbon vs Kimi K2.5
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 6,100+ subscribers.
Last updated October 28, 2026