Long documents
Prompts that approach the documented context limit
Gemini 3 Flash
Gemini 3 Flash has the larger documented context window.
Confidence: documented
Gemini 3 Flash vs Kimi K2.5
Updated September 29, 2026. Rank says Gemini 3 Flash is ahead. Price, access, and your workload can each overturn that. 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
Decision reading
Gemini 3 Flash has the higher public score estimate, 55.5 versus 52.85, but the 90% score intervals overlap. Treat that as a lead, not a settled winner. 5 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
- Shared results
- 5
- Gemini 3 Flash only
- 6
- Kimi K2.5 only
- 40
- Like-for-like categories
- 0 / 8
Supported: Gemini 3 Flash and Kimi K2.5How 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.
Chat turn cost
1K fresh input + 500 output tokens
Gemini 3 Flash
Gemini 3 Flash has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-ratesCache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Gemini 3 Flash
Gemini 3 Flash 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
Gemini 3 Flash
Gemini 3 Flash 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
Kimi K2.5 is scored on Estimated evidence for agentic, so the reading is directional rather than like-for-like.
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.
36.4Gemini 3 Flash38.8Kimi K2.5
Directional only · BenchAlign v5.7
Kimi K2.5 scores higher, but at least one score rests on Estimated evidence or a different benchmark set. Directional only, no winner.
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.
6 Gemini 3 Flash only
5 Shared
40 Kimi K2.5 only
4 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.
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.
FrontierMath v2 (Tiers 1-3)Math
Normalized gap 7.7Gemini 3 Flash:35.640%Kimi K2.5:27.900%FrontierMath v2 (Tier 4)Math
Normalized gap 0.0Gemini 3 Flash:4.167%Kimi K2.5:4.200%
Bars run 0–100 on each benchmark’s normalized display scale
Category results, on a stated basis
Each row shows the public-lane category score for both models: the BenchAlign v5.7 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
Directional only- Gemini 3 Flash
- 31.7
- Supported · #76/117
- Kimi K2.5
- 36.5
- Estimated · #61/117
- Basis
- BenchAlign v5.7 lane · 4 vs 14 public rows
- Reading
- Directional only
Coding
Directional only- Gemini 3 Flash
- 36.4
- Supported · #74/143
- Kimi K2.5
- 38.8
- Estimated · #66/143
- Basis
- BenchAlign v5.7 lane · 3 vs 8 public rows
- Reading
- Directional only
Knowledge
Directional only- Gemini 3 Flash
- 54.3
- Supported · #47/169
- Kimi K2.5
- 47.8
- Estimated · #71/169
- Basis
- BenchAlign v5.7 lane · 2 vs 6 public rows
- Reading
- Directional only
Instruction following
Directional only- Gemini 3 Flash
- 64.7
- #70/124
- Kimi K2.5
- 84.4
- #41/124
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Directional only
Reasoning
Not comparable- Gemini 3 Flash
- 60.3
- Unranked · 2 rankable rows
- Kimi K2.5
- 55.2
- Unranked · 3 rankable rows
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Multimodal
Not comparable- Gemini 3 Flash
- 76.9
- Unranked · 1 rankable row
- Kimi K2.5
- 66.8
- #25/50
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Gemini 3 Flash
- Not ranked
- Kimi K2.5
- 38.2
- #8/12
- Basis
- Provisional lane · 0 vs 1 weighted rows
- Reading
- Not comparable
Math
Not comparable- Gemini 3 Flash
- 50.0
- Unranked · 2 rankable rows
- Kimi K2.5
- 62.1
- #4/7
- Basis
- Provisional lane · 2 vs 4 weighted rows
- Reading
- Not comparable
| Category | Gemini 3 Flash | Kimi K2.5 | Basis | Reading |
|---|---|---|---|---|
| Agentic | 31.7Supported · #76/117 | 36.5Estimated · #61/117 | Directional onlyBenchAlign v5.7 lane · 4 vs 14 public rows | Directional only |
| Coding | 36.4Supported · #74/143 | 38.8Estimated · #66/143 | Directional onlyBenchAlign v5.7 lane · 3 vs 8 public rows | Directional only |
| Knowledge | 54.3Supported · #47/169 | 47.8Estimated · #71/169 | Directional onlyBenchAlign v5.7 lane · 2 vs 6 public rows | Directional only |
| Instruction following | 64.7#70/124 | 84.4#41/124 | Directional onlyProvisional lane · 0 vs 0 weighted rows | Directional only |
| Reasoning | 60.3Unranked · 2 rankable rows | 55.2Unranked · 3 rankable rows | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Multimodal | 76.9Unranked · 1 rankable row | 66.8#25/50 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Multilingual | Not ranked | 38.2#8/12 | Not comparableProvisional lane · 0 vs 1 weighted rows | Not comparable |
| Math | 50.0Unranked · 2 rankable rows | 62.1#4/7 | Not comparableProvisional lane · 2 vs 4 weighted rows | Not comparable |
Ranks count the models scored in each category’s lane, so the agentic, coding, and knowledge denominators (BenchAlign v5.7) 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 3 Flash
- $0.002
- Fits in one request
- Kimi K2.5
- $0.0021
- Fits in one request
Gemini 3 Flash has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Gemini 3 Flash
- $0.034
- Fits in one request
- Kimi K2.5
- $0.039
- Fits in one request
Gemini 3 Flash 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 3 Flash
- $0.05
- Fits in one request
- Kimi K2.5
- $0.162
- Fits in one request
- Cached input priced at the published list-input rate
Gemini 3 Flash 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 3 FlashKimi K2.5
Context window
Maximum documented context; output-token limits may be lower.
Gemini 3 Flash
Kimi K2.5
256K
API model ID
Gemini 3 Flash
gemini-3-flash-preview
Google Gemini API pricingKimi 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 3 Flash
$0.05 per 1M cached input tokens
Google Gemini API pricingKimi K2.5
Not published
Documented inputs
Gemini 3 Flash
Not sourced
Kimi K2.5
Not sourced
Documented outputs
Gemini 3 Flash
Not sourced
Kimi K2.5
Not sourced
Provider availability
Gemini 3 Flash
Not sourced
Kimi K2.5
Not sourced
Reasoning profile
Gemini 3 Flash
Non-Reasoning
Kimi K2.5
Non-Reasoning
Weight access
Gemini 3 Flash
Proprietary
Kimi K2.5
Open Weight
License
Gemini 3 Flash
Proprietary
Kimi K2.5
Open Weight
Release date
Gemini 3 Flash
2025-12-01
Kimi K2.5
2026-02-01
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
- Gemini 3 Flash has the higher public score estimate, 55.5 versus 52.85, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.034 vs $0.039. Cache-heavy agent loop: $0.05 vs $0.162.
- Context tradeoff
- Gemini 3 Flash has the larger documented window (1M).
Run the same representative tasks against both endpoints before changing production traffic.
Questions
Which is better, Gemini 3 Flash or Kimi K2.5?
Gemini 3 Flash has the higher public score estimate, 55.5 versus 52.85, but the 90% score intervals overlap. The higher estimate is not a decisive winner because the uncertainty ranges overlap.
Which is better for coding, Gemini 3 Flash or Kimi K2.5?
Kimi K2.5 scores higher for coding on the public lane, 38.8 to 36.4. 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 3 Flash or Kimi K2.5?
Kimi K2.5 scores higher for agentic tasks on the public lane, 36.5 to 31.7. Kimi K2.5 is scored on Estimated evidence for agentic tasks, so the reading is directional rather than like-for-like. BenchLM does not name a winner for agentic tasks from a directional reading; read the shared benchmark rows directly and test the models on the same task set.
Which costs less, Gemini 3 Flash or Kimi K2.5?
For the stated presets, chat costs $0.002 on Gemini 3 Flash and $0.0021 on Kimi K2.5; repository review costs $0.034 and $0.039; the cache-heavy agent loop costs $0.05 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 3 Flash or Kimi K2.5?
Gemini 3 Flash has the larger documented context window: 1M, compared with 256K.
Self-host vs API cost
Estimates at 50,000 req/day · 1000 tokens/req average.
Gemini 3 Flash
API / mo$2,625
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.5
API / mo$2,700
Self-host / mo$5,221
Break-even132M/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 evidence51 rows
Agentic
Claw-Eval
Shared sourceGemini 3 Flash49.2%Kimi K2.552.3%Kimi K2.5 leads this result
Gert Labs
Shared sourceGemini 3 Flash56.63%Kimi K2.545.88%Gemini 3 Flash leads this result
JobBench
Shared sourceGemini 3 Flash11.4%Kimi K2.58.7%Gemini 3 Flash leads this result
Terminal-Bench 2.1 (Vals)
Gemini 3 Flash53.9%SourceKimi K2.5—Not directly comparable
Terminal-Bench 2.0
Gemini 3 Flash—Kimi K2.550.8%SourceNot directly comparable
BrowseComp
Gemini 3 Flash—Kimi K2.560.6%SourceNot directly comparable
QwenClawBench
Gemini 3 Flash—Kimi K2.554.3%SourceNot directly comparable
τ³-bench results
Gemini 3 Flash—Kimi K2.565.7%SourceNot directly comparable
DeepSearchQA
Gemini 3 Flash—Kimi K2.577.1%SourceNot directly comparable
DeepPlanning
Gemini 3 Flash—Kimi K2.514.4%SourceNot directly comparable
Toolathlon
Gemini 3 Flash—Kimi K2.527.8%SourceNot directly comparable
MCP Atlas
Gemini 3 Flash—Kimi K2.529.5%SourceNot directly comparable
MCP-Tasks
Gemini 3 Flash—Kimi K2.559.1%SourceNot directly comparable
WideResearch
Gemini 3 Flash—Kimi K2.572.7%SourceNot directly comparable
ResearchClawBench
Gemini 3 Flash—Kimi K2.514.0%SourceNot directly comparable
Coding
Vibe Code Bench
Gemini 3 Flash20.20%SourceKimi K2.5—Not directly comparable
LiveCodeBench (Vals)
Gemini 3 Flash85.6%SourceKimi K2.5—Not directly comparable
SWE-bench (Vals)
Gemini 3 Flash75.0%SourceKimi K2.5—Not directly comparable
SWE-bench Verified
Gemini 3 Flash—Kimi K2.576.8%SourceNot directly comparable
SWE-bench Verified*
Gemini 3 Flash—Kimi K2.570.8%SourceNot directly comparable
LiveCodeBench v6
Gemini 3 Flash—Kimi K2.585.0%SourceNot directly comparable
SWE-bench Pro
Gemini 3 Flash—Kimi K2.550.7%SourceNot directly comparable
SWE Multilingual
Gemini 3 Flash—Kimi K2.573%SourceNot directly comparable
SWE-Rebench
Gemini 3 Flash—Kimi K2.558.5%SourceNot directly comparable
React Native Evals
Gemini 3 Flash—Kimi K2.577.2%SourceNot directly comparable
SciCode
Gemini 3 Flash—Kimi K2.548.7%SourceNot directly comparable
Reasoning
LongBench v2
Gemini 3 Flash—Kimi K2.561%SourceNot directly comparable
Multimodal
Knowledge
GPQA Diamond (Vals)
Gemini 3 Flash87.9%SourceKimi K2.5—Not directly comparable
MMLU-Pro (Vals)
Gemini 3 Flash88.6%SourceKimi K2.5—Not directly comparable
GPQA
Gemini 3 Flash—Kimi K2.587.6%SourceNot directly comparable
GPQA-D
Gemini 3 Flash—Kimi K2.587.6%SourceNot directly comparable
SuperGPQA
Gemini 3 Flash—Kimi K2.569.2%SourceNot directly comparable
MMLU-Pro
Gemini 3 Flash—Kimi K2.587.1%SourceNot directly comparable
MMLU-Pro (Arcee)
Gemini 3 Flash—Kimi K2.587.1%SourceNot directly comparable
HLE
Gemini 3 Flash—Kimi K2.530.1%SourceNot directly comparable
Multilingual
Instruction following
IFEval
Gemini 3 Flash—Kimi K2.593.9%SourceNot directly comparable
Math
FrontierMath v2 (Tiers 1-3)
Shared sourceGemini 3 Flash35.640%Kimi K2.527.900%Gemini 3 Flash leads this result
FrontierMath v2 (Tier 4)
Shared sourceGemini 3 Flash4.167%Kimi K2.54.200%Kimi K2.5 leads this result
AIME 2025
Gemini 3 Flash—Kimi K2.596.1%SourceNot directly comparable
AIME26
Gemini 3 Flash—Kimi K2.595.8%SourceNot directly comparable
AIME25 (Arcee)
Gemini 3 Flash—Kimi K2.596.3%SourceNot directly comparable
HMMT Feb 2025
Gemini 3 Flash—Kimi K2.595.4%SourceNot directly comparable
HMMT Nov 2025
Gemini 3 Flash—Kimi K2.591.1%SourceNot directly comparable
HMMT Feb 2026
Gemini 3 Flash—Kimi K2.587.1%SourceNot directly comparable
MMAnswerBench
Gemini 3 Flash—Kimi K2.581.8%SourceNot directly comparable
51 public results · 5 shared
Watch Gemini 3 Flash vs Kimi K2.5
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.
Last updated September 29, 2026