Coding
Like-for-like- Gemini 3.5 Flash-Lite
- 43.6
- Supported · #121/183
- Kimi K2.6
- 51.6
- Supported · #62/183
- Basis
- BenchAlign lane · 4 vs 10 public rows
- Reading
- Kimi K2.6 leads · intervals overlap
Gemini 3.5 Flash-Lite vs Kimi K2.6
Updated September 4, 2026. Public scores include evidence status and uncertainty. They are not guarantees for a specific workload.
Decision reading
Kimi K2.6 has the higher public score estimate, 65.3 versus 60.5, but the 90% score intervals overlap. Treat that as a lead, not a settled winner.
9 results are shared. Category rows resting on Estimated evidence or different benchmark sets are marked directional and do not name a winner.
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.
Coding work
Code generation, repair, and software-engineering tasks
Kimi K2.6
Kimi K2.6 leads on the public coding lane, 51.6 to 43.6, with Supported evidence for both models, although the 90% intervals overlap.
Confidence: limited
Long documents
Prompts that approach the documented context limit
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite has the larger documented context window.
Confidence: documented
Chat turn cost
1K fresh input + 500 output tokens
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Show secondary and unsupported calls
Cache-heavy agent loop cost
200K cached + 20K fresh input + 10K output tokens
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite 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
Repository review cost
50K fresh input + 3K output tokens
Gemini 3.5 Flash-Lite
Gemini 3.5 Flash-Lite has the lower estimated token cost for this stated workload. Costs use the listed standard API rates.
Confidence: listed-rates
Agentic work
Tool use, computer use, and multi-step task completion
Not enough matched evidence
Gemini 3.5 Flash-Lite and Kimi K2.6 are scored on Estimated evidence for agentic, so the reading is 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.
1 Gemini 3.5 Flash-Lite only
9 Shared
28 Kimi K2.6 only
- Shared results
- 9
- Gemini 3.5 Flash-Lite only
- 1
- Kimi K2.6 only
- 28
- Like-for-like categories
- 2 / 8
2 categories rest 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.
Knowledge
Like-for-like- Gemini 3.5 Flash-Lite
- 53.0
- Supported · #71/181
- Kimi K2.6
- 61.7
- Supported · #34/181
- Basis
- BenchAlign lane · 2 vs 5 public rows
- Reading
- Kimi K2.6 leads · intervals overlap
Agentic
Directional only- Gemini 3.5 Flash-Lite
- 43.4
- Estimated · #104/151
- Kimi K2.6
- 46.1
- Estimated · #88/151
- Basis
- BenchAlign lane · 3 vs 12 public rows
- Reading
- Directional only
Multimodal
Directional only- Gemini 3.5 Flash-Lite
- 76.1
- #16/48
- Kimi K2.6
- 63.2
- #26/48
- Basis
- Provisional lane · 0 vs 2 weighted rows
- Reading
- Directional only
Reasoning
Not comparable- Gemini 3.5 Flash-Lite
- 60.8
- Unranked · 3 rankable rows
- Kimi K2.6
- Not ranked
- Basis
- Provisional lane · 1 vs 0 weighted rows
- Reading
- Not comparable
Math
Not comparable- Gemini 3.5 Flash-Lite
- Not ranked
- Kimi K2.6
- 71.3
- #1/7
- Basis
- Provisional lane · 0 vs 4 weighted rows
- Reading
- Not comparable
Multilingual
Not comparable- Gemini 3.5 Flash-Lite
- Not ranked
- Kimi K2.6
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
Instruction following
Not comparable- Gemini 3.5 Flash-Lite
- Not ranked
- Kimi K2.6
- Not ranked
- Basis
- Provisional lane · 0 vs 0 weighted rows
- Reading
- Not comparable
| Category | Gemini 3.5 Flash-Lite | Kimi K2.6 | Basis | Reading |
|---|---|---|---|---|
| Coding | 43.6Supported · #121/183 | 51.6Supported · #62/183 | Like-for-likeBenchAlign lane · 4 vs 10 public rows | Kimi K2.6 leads · intervals overlap |
| Knowledge | 53.0Supported · #71/181 | 61.7Supported · #34/181 | Like-for-likeBenchAlign lane · 2 vs 5 public rows | Kimi K2.6 leads · intervals overlap |
| Agentic | 43.4Estimated · #104/151 | 46.1Estimated · #88/151 | Directional onlyBenchAlign lane · 3 vs 12 public rows | Directional only |
| Multimodal | 76.1#16/48 | 63.2#26/48 | Directional onlyProvisional lane · 0 vs 2 weighted rows | Directional only |
| Reasoning | 60.8Unranked · 3 rankable rows | Not ranked | Not comparableProvisional lane · 1 vs 0 weighted rows | Not comparable |
| Math | Not ranked | 71.3#1/7 | Not comparableProvisional lane · 0 vs 4 weighted 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 |
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.
Terminal-Bench 2.0
Agentic
Gemini 3.5 Flash-Lite: 54%Kimi K2.6: 66.7%Normalized gap 12.7Gemini 3.5 Flash-Lite source Kimi K2.6 sourceLiveCodeBench (Vals)
Coding
Gemini 3.5 Flash-Lite: 79.0%Kimi K2.6: 86.8%Normalized gap 7.8Gemini 3.5 Flash-Lite source Kimi K2.6 sourceSWE-bench Pro
Coding
Gemini 3.5 Flash-Lite: 54.2%Kimi K2.6: 58.6%Normalized gap 4.4Gemini 3.5 Flash-Lite source Kimi K2.6 sourceMMLU-Pro (Vals)
Knowledge
Gemini 3.5 Flash-Lite: 85.8%Kimi K2.6: 87.6%Normalized gap 1.8Gemini 3.5 Flash-Lite source Kimi K2.6 sourceOSWorld-Verified
Agentic
Gemini 3.5 Flash-Lite: 74%Kimi K2.6: 73.1%Normalized gap 0.9Gemini 3.5 Flash-Lite source Kimi K2.6 source
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 3.5 Flash-Lite
- $0.00155
- Fits in one request
- Kimi K2.6
- $0.00295
- Fits in one request
Gemini 3.5 Flash-Lite has the lower modeled cost
Costs use the listed standard API rates.
Repository review
50K fresh input + 3K output tokens
- Gemini 3.5 Flash-Lite
- $0.0225
- Fits in one request
- Kimi K2.6
- $0.0595
- Fits in one request
Gemini 3.5 Flash-Lite 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 3.5 Flash-Lite
- $0.037
- Fits in one request
- Kimi K2.6
- $0.249
- Fits in one request
- Cached input priced at the published list-input rate
Gemini 3.5 Flash-Lite 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.
SpecificationGemini 3.5 Flash-LiteKimi K2.6
Context window
Maximum documented context; output-token limits may be lower.
Gemini 3.5 Flash-Lite
Kimi K2.6
256K
API model ID
Gemini 3.5 Flash-Lite
gemini-3.5-flash-lite
Google Gemini 3.5 Flash-Lite model documentationKimi K2.6
Not sourced
Cached-input rate
A missing cached-input rate falls back to the listed input rate only in the stated workload estimate.
Gemini 3.5 Flash-Lite
$0.03 per 1M cached input tokens
Google Gemini API pricingKimi K2.6
Not published
Documented inputs
Gemini 3.5 Flash-Lite
text, image, video, audio, pdf
Google Gemini 3.5 Flash-Lite model documentationKimi K2.6
Not sourced
Documented outputs
Gemini 3.5 Flash-Lite
Kimi K2.6
Not sourced
Provider availability
Gemini 3.5 Flash-Lite
Generally Available · Gemini API, Google AI Studio
Google latest Gemini model guideKimi K2.6
Not sourced
Reasoning profile
Gemini 3.5 Flash-Lite
Reasoning
Kimi K2.6
Reasoning
Weight access
Gemini 3.5 Flash-Lite
Proprietary
Kimi K2.6
Open Weight
License
Gemini 3.5 Flash-Lite
Proprietary
Kimi K2.6
Open Weight
Release date
Gemini 3.5 Flash-Lite
2026-07-21
Kimi K2.6
2026-04-20
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
- Kimi K2.6 has the higher public score estimate, 65.3 versus 60.5, but the 90% score intervals overlap.
- Workload cost
- Repository review: $0.0225 vs $0.0595. Cache-heavy agent loop: $0.037 vs $0.249.
- Context tradeoff
- Gemini 3.5 Flash-Lite 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.
Gemini 3.5 Flash-Lite
API / mo$2,100
Self-host / moNot listed
Break-even—
Proprietary model — self-hosting not applicable.
Kimi K2.6
API / mo$3,713
Self-host / mo$18,221
Break-even326M/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 evidence38 rows
Agentic
Terminal-Bench 2.0
Gemini 3.5 Flash-Lite54%SourceKimi K2.666.7%SourceKimi K2.6 leads this result
OSWorld-Verified
Gemini 3.5 Flash-Lite74%SourceKimi K2.673.1%SourceGemini 3.5 Flash-Lite leads this result
Terminal-Bench 2.1 (Vals)
Gemini 3.5 Flash-Lite50.2%SourceKimi K2.653.6%SourceKimi K2.6 leads this result
BrowseComp
Gemini 3.5 Flash-Lite—Kimi K2.683.2%SourceNot directly comparable
Toolathlon
Gemini 3.5 Flash-Lite—Kimi K2.650%SourceNot directly comparable
MCP Atlas
Gemini 3.5 Flash-Lite—Kimi K2.655.9%SourceNot directly comparable
Claw-Eval
Gemini 3.5 Flash-Lite—Kimi K2.662.3%SourceNot directly comparable
DeepSearchQA
Gemini 3.5 Flash-Lite—Kimi K2.692.5%SourceNot directly comparable
WideResearch
Gemini 3.5 Flash-Lite—Kimi K2.680.8%SourceNot directly comparable
Gert Labs
Gemini 3.5 Flash-Lite—Kimi K2.656.82%SourceNot directly comparable
ResearchClawBench
Gemini 3.5 Flash-Lite—Kimi K2.618.0%SourceNot directly comparable
OSWorld 2.0
Gemini 3.5 Flash-Lite—Kimi K2.64.6%SourceNot directly comparable
Coding
Terminal-Bench 2.0
Gemini 3.5 Flash-Lite54.0%SourceKimi K2.666.7%SourceKimi K2.6 leads this result
SWE-bench Pro
Gemini 3.5 Flash-Lite54.2%SourceKimi K2.658.6%SourceKimi K2.6 leads this result
LiveCodeBench (Vals)
Gemini 3.5 Flash-Lite79.0%SourceKimi K2.686.8%SourceKimi K2.6 leads this result
SWE-bench (Vals)
Gemini 3.5 Flash-Lite75.0%SourceKimi K2.676.2%SourceKimi K2.6 leads this result
SWE-bench Verified
Gemini 3.5 Flash-Lite—Kimi K2.680.2%SourceNot directly comparable
LiveCodeBench v6
Gemini 3.5 Flash-Lite—Kimi K2.689.6%SourceNot directly comparable
SWE Multilingual
Gemini 3.5 Flash-Lite—Kimi K2.676.7%SourceNot directly comparable
SciCode
Gemini 3.5 Flash-Lite—Kimi K2.652.2%SourceNot directly comparable
Vibe Code Bench
Gemini 3.5 Flash-Lite—Kimi K2.637.89%SourceNot directly comparable
cursorBench31
Gemini 3.5 Flash-Lite—Kimi K2.647.6%SourceNot directly comparable
Reasoning
MRCRv2
Gemini 3.5 Flash-Lite72.2%SourceKimi K2.6—Not directly comparable
Knowledge
GPQA Diamond (Vals)
Gemini 3.5 Flash-Lite83.8%SourceKimi K2.689.1%SourceKimi K2.6 leads this result
MMLU-Pro (Vals)
Gemini 3.5 Flash-Lite85.8%SourceKimi K2.687.6%SourceKimi K2.6 leads this result
GPQA
Gemini 3.5 Flash-Lite—Kimi K2.690.5%SourceNot directly comparable
GPQA-D
Gemini 3.5 Flash-Lite—Kimi K2.690.5%SourceNot directly comparable
HLE
Gemini 3.5 Flash-Lite—Kimi K2.634.7%SourceNot directly comparable
Math
AIME26
Gemini 3.5 Flash-Lite—Kimi K2.696.4%SourceNot directly comparable
HMMT Feb 2026
Gemini 3.5 Flash-Lite—Kimi K2.692.7%SourceNot directly comparable
MMAnswerBench
Gemini 3.5 Flash-Lite—Kimi K2.686.0%SourceNot directly comparable
FrontierMath v2 (Tiers 1-3)
Gemini 3.5 Flash-Lite—Kimi K2.638.966%SourceNot directly comparable
FrontierMath v2 (Tier 4)
Gemini 3.5 Flash-Lite—Kimi K2.614.580%SourceNot directly comparable
Multimodal
MMMU-Pro
Gemini 3.5 Flash-Lite—Kimi K2.679.4%SourceNot directly comparable
MMMU-Pro w/ Python
Gemini 3.5 Flash-Lite—Kimi K2.680.1%SourceNot directly comparable
CharXiv
Gemini 3.5 Flash-Lite—Kimi K2.680.4%SourceNot directly comparable
MathVision
Gemini 3.5 Flash-Lite—Kimi K2.687.4%SourceNot directly comparable
V*
Gemini 3.5 Flash-Lite—Kimi K2.696.9%SourceNot directly comparable
Frequently asked questions
Which is better, Gemini 3.5 Flash-Lite or Kimi K2.6?
Kimi K2.6 has the higher public score estimate, 65.3 versus 60.5, 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 3.5 Flash-Lite or Kimi K2.6?
Kimi K2.6 leads the public coding lane, 51.6 to 43.6, with Supported evidence for both models, although the 90% intervals overlap.
Which is better for agentic tasks, Gemini 3.5 Flash-Lite or Kimi K2.6?
Kimi K2.6 scores higher for agentic tasks on the public lane, 46.1 to 43.4. Gemini 3.5 Flash-Lite and Kimi K2.6 are 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 3.5 Flash-Lite or Kimi K2.6?
For the stated presets, chat costs $0.00155 on Gemini 3.5 Flash-Lite and $0.00295 on Kimi K2.6; repository review costs $0.0225 and $0.0595; the cache-heavy agent loop costs $0.037 and $0.249. Kimi K2.6 has no published cached-input rate, so cached tokens use its listed input rate.
Which has the larger context window, Gemini 3.5 Flash-Lite or Kimi K2.6?
Gemini 3.5 Flash-Lite has the larger documented context window: 1M, compared with 256K.
Related comparisons
Last updated September 4, 2026
Watch Gemini 3.5 Flash-Lite vs Kimi K2.6
One weekly email when material rank, price, or benchmark evidence changes make this matchup worth revisiting.
Read a sample issueJoin 2,000+ readers.