Is Gemini 3.5 Flash better than Kimi K3?

Gemini 3.5 Flash (high) and Kimi K3 (max) are neck and neck on quality (63.1 vs 64.8, within the ±4.3 margin). The BenchLeader Index combines every independent quality benchmark; the two are within the margin of error as of 2026-10-25, so check the category scores for your use.

Which is cheaper, Gemini 3.5 Flash or Kimi K3, and will that change?

Gemini 3.5 Flash is cheaper today: $3.38 against $6.00 per million tokens, blended at three input tokens per output token. No price change has been announced for either.

Is Gemini 3.5 Flash better than Kimi K3 for coding?

Kimi K3 scores higher in coding (65 vs 60 on the category index, where 50 is average). On Arena WebDev, the top shared coding benchmark, it is 1499 for Gemini 3.5 Flash against 1655 for Kimi K3.

Is Gemini 3.5 Flash better than Kimi K3 for agents and tool use?

Gemini 3.5 Flash scores higher in agentic tasks (63 vs 59 on the category index, where 50 is average). On MCP Atlas, the top shared agents & tools benchmark, it is 83.6% for Gemini 3.5 Flash against 82.3% for Kimi K3.

Which is better for long documents?

Kimi K3 scores higher on long-context benchmarks (70 vs 62). Both accept 1,048,576 tokens of context.

Which has lower latency?

Gemini 3.5 Flash (high) answers first: 14.2 s against 52.6 s to the first answer token, as Artificial Analysis measures it. Reasoning effort drives most of this: a higher setting thinks longer before it answers.

Which is faster, Gemini 3.5 Flash or Kimi K3?

Gemini 3.5 Flash streams faster: 222 against 41 output tokens per second.

When was Kimi K3 released?

Kimi K3 was released on 16 Jul 2026.

Gemini 3.5 Flash vs Kimi K3

Full profiles: Gemini 3.5 Flash (Google) and Kimi K3 (Moonshot AI): every benchmark, price and rank.

Updated 25 Oct 2026 · Data from 10 sources

As of 25 Oct 2026, Gemini 3.5 Flash (high) and Kimi K3 (max) are neck and neck on quality (63.1 vs 64.8 on the BenchLeader Index, inside the ±4.3 margin of error), and Gemini 3.5 Flash (high) costs 1.8× less, which makes it the better buy for most workloads. Pick Gemini 3.5 Flash for agents & tools (63 vs 59); pick Kimi K3 for long context (70 vs 62).

Gemini 3.5 Flash benchmarks high Google

Kimi K3 benchmarks max Moonshot AI Open in the comparison tool

Which to pick for…

The pick's numbers come first. Category scores on the BenchLeader scale (50 is the average model, 15 points one standard deviation). High confidence means a lead of half a standard deviation or more; under one point is called either.

Use case	Pick	Why	Confidence
Coding	Kimi K3	65 vs 60 category score; Arena WebDev 1655 vs 1499	Medium
Agents & tool use	Gemini 3.5 Flash	63 vs 59 category score; MCP Atlas 83.6% vs 82.3%	Medium
Maths	Kimi K3	63 vs 58 category score; FrontierMath Tiers 1–3 72.2% vs 62.8%	Medium
Reasoning	Kimi K3	66 vs 61 category score; GPQA Diamond 93.1% vs 92.8%	Medium
Knowledge	Gemini 3.5 Flash	65 vs 63 category score; SimpleQA Verified 66.2% vs 50.6%	Medium
Long documents	Kimi K3	70 vs 62 category score; AA-LCR 88.7% vs 73.3%	High
Images and other media	Gemini 3.5 Flash	67 vs 64 category score; MMMU-Pro 84.3% vs 80.5%	Medium
Chat quality (human preference)	Kimi K3	68 vs 67 on Arena votes	Medium
Lowest cost	Gemini 3.5 Flash	$3.38 vs $6.00 per 1M blended; agentic coding session $0.126 vs $0.240	–
Fastest first answer	Gemini 3.5 Flash	14.2 s (Gemini 3.5 Flash (high)) vs 52.6 s (Kimi K3 (max)) to the first answer token	–
High volume on a budget	Neither flagship	A smaller model is usually the better fit: Cheapest good LLMs, Fastest LLMs	–

Key differences

Gemini 3.5 Flash (high) and Kimi K3 (max) are neck and neck on quality (63.1 vs 64.8, within the ±4.3 margin).
Gemini 3.5 Flash (high) is stronger in agents & tools, instruction following, knowledge, multimodal.
Kimi K3 (max) is stronger in coding, composite, human preference, long context, maths, reasoning.
Gemini 3.5 Flash (high) is 1.8× cheaper ($3.38 vs $6.00 per 1M blended).
Gemini 3.5 Flash (high) streams 5.5× faster (222 vs 41 tokens per second).

At a glance

Spec	Gemini 3.5 Flash	Kimi K3
Developer	Google	Moonshot AI
Released	19 May 2026	16 Jul 2026
Context window (tokens)	1,048,576	1,048,576
Compared at	Gemini 3.5 Flash (high)	Kimi K3 (max)
Open weights	No	Yes
Price in / out / cached input per 1M	$1.50 / $9.00 / $0.150	$3.00 / $15.00 / $0.300

Category by category

Category scores on the BenchLeader scale: 50 is the average evaluated model, 15 points is one standard deviation. A gap under one point is called level.

Category	Gemini 3.5 Flash (high)	Kimi K3 (max)	Leader
Reasoning	61.3	66.4	Kimi K3
Coding	59.8	64.6	Kimi K3
Agents & tools	63.4	59.4	Gemini 3.5 Flash
Maths	58.3	63.4	Kimi K3
Knowledge	65.2	62.8	Gemini 3.5 Flash
Instruction following	74.3	–	not comparable
Human preference	66.8	68.1	Kimi K3
Multimodal	66.6	63.6	Gemini 3.5 Flash
Long context	61.6	69.6	Kimi K3
Composite	65.9	78.3	Kimi K3

Where sources disagree

Across the 59 benchmarks both have, Gemini 3.5 Flash leads 19, Kimi K3 leads 40. Every overall index that rates both puts Kimi K3 ahead.

Overall index	Gemini 3.5 Flash (high)	Kimi K3 (max)	Leader
BenchLeader Index	63.1	64.8	neck and neck
AA Intelligence Index v4.3.2	32.6	43.6	Kimi K3
Arena Text	1477	1487	Kimi K3

Each publisher runs its own tasks, harnesses and effort settings, so their overall verdicts can differ; BenchLeader reconciles them on one scale. † measured at another effort setting of the same model (the publisher did not test the one compared here).

Price, speed and latency

List prices blended at three input tokens per output token. Per-task prices use the token counts on the methodology page; suite cost is what Artificial Analysis paid to run its whole Intelligence Index.

Measure	Gemini 3.5 Flash (high)	Kimi K3 (max)
Blended price per 1M tokens	$3.38	$6.00
Input / output per 1M	$1.50 / $9.00	$3.00 / $15.00
Cached input per 1M	$0.150	$0.300
Price per task: chat reply	$0.0033	$0.0057
Price per task: summarize a 30-page report	$0.023	$0.045
Price per task: code edit	$0.022	$0.041
Price per task: agentic coding session	$0.126	$0.240
Price per task: structured extraction	$0.0048	$0.0090
Cost to run a full benchmark suite	$3.69	$2.00
Output speed	222 tok/s	41 tok/s
Time to first answer	14.2 s	52.6 s
Time to a complete answer	16.4 s	65.0 s

Per token is not per task: on Artificial Analysis' full suite Gemini 3.5 Flash (high) cost $3.69 and Kimi K3 (max) $2.00, because Gemini 3.5 Flash (high) used about 3.3× as many tokens for the same work, which a per-token price hides.

Each model is compared at its best-ranked reasoning-effort setting. Gemini 3.5 Flash (high) scores 63.1 and takes 14 s to its first answer token; Gemini 3.5 Flash (medium) scores 62.8 in 12 s, Gemini 3.5 Flash (minimal) scores 56.4 in 1 s. Kimi K3 (max) scores 64.8 and takes 53 s to its first answer token; Kimi K3 scores 58.8 in 1 s, Kimi K3 (low) scores 58.2 in 57 s. Higher effort means more thinking before the answer: better scores, longer waits and more tokens billed. How reasoning effort changes scores and cost.

Every benchmark

144 benchmarks, grouped by category; each name links to the board and its leaders. † measured at another effort setting of the same model, shown for reference and not counted as a win.

Metric	Publisher	Gemini 3.5 Flash (high)	Kimi K3 (max)
BenchLeader Index	BenchLeader	63.1	64.8
Agents & tools score	BenchLeader	63.4	59.4
Coding score	BenchLeader	59.8	64.6
Composite score	BenchLeader	65.9	78.3
Human preference score	BenchLeader	66.8	68.1
Instruction following score	BenchLeader	74.3	–
Knowledge score	BenchLeader	65.2	62.8
Long context score	BenchLeader	61.6	69.6
Maths score	BenchLeader	58.3	63.4
Multimodal score	BenchLeader	66.6	63.6
Reasoning score	BenchLeader	61.3	66.4
Blended price $/M	BenchLeader	$3.38	$6.00
Output speed	BenchLeader	222 tok/s	41 tok/s
Time to first answer	BenchLeader	14.2 s	52.6 s
Context window	BenchLeader	1,048,576 tokens	1,048,576 tokens
Reasoning
GPQA Diamond	Epoch AI Benchmarking Hub	92.8%	93.1%
SimpleBench	SimpleBench	76.7%†	60.7%
Arena Hard Prompts	Arena (LMArena)	1495	1516
LiveBench 2.0 Reasoning	LiveBench	–	90.7%†
LiveBench 1.0 Reasoning	LiveBench	82.0%	90.7%†
GPQA Diamond (AA)	Artificial Analysis	92.2%	93.5%
Humanity's Last Exam (AA)	Artificial Analysis	42.7%	46.9%
GPQA Diamond (Vals)	Vals AI	92.7%	92.9%†
EnigmaEval	Scale AI SEAL	25.4%	–
ARC-AGI-1	ARC Prize	92.5%	94.5%
ARC-AGI-2	ARC Prize	72.1%	60.4%
CritPt	Artificial Analysis	13.1%	23.4%
Chess Puzzles	Epoch AI Benchmarking Hub	45.0%	36.0%
Mystery Game Puzzles	Epoch AI Benchmarking Hub	32.0%	26.0%
LMCA	Epoch AI Benchmarking Hub	47.1%	52.7%
DTBench	Epoch AI Benchmarking Hub	94.7%	91.2%
ForecastBench	ForecastBench	64.3%†	64.6%
Coding
SWE-bench Verified (Epoch)	Epoch AI Benchmarking Hub	79.3%	–
SciCode	SciCode	53.9%	59.5%
WeirdML	WeirdML	62.6%	82.6%
FrontierCode	Cognition	–	44.2%†
Arena Coding	Arena (LMArena)	1508	1539
Arena WebDev	Arena (LMArena)	1499	1655
LiveBench 2.0 Coding	LiveBench	–	59.3%†
LiveBench 1.0 Coding	LiveBench	78.2%	81.5%†

Show all 144 benchmarks

Metric	Publisher	Gemini 3.5 Flash (high)	Kimi K3 (max)
Coding
LiveCodeBench	Vals AI	87.6%	87.2%
IOI	Vals AI	–	48.9%
SWE-bench (Vals)	Vals AI	78.8%	93.4%†
Code Migration	Vals AI	26.8%	16.1%
Code Migration v1.1	Vals AI	–	28.6%
ProgramBench	Vals AI	0.0%	2.0%†
SWE-bench Multimodal	Vals AI	–	44.2%
Vibe Code Bench 1-100	Vals AI	–	18.2%
Vibe Code Bench v1.1	Vals AI	48.7%	85.0%
Surface Evolver Bench	Epoch AI Benchmarking Hub	58.1%†	93.0%
DeepSWE v1.1	Epoch AI Benchmarking Hub	36.1%	68.5%
ALE-Bench	Epoch AI Benchmarking Hub	911.0	1524.5
FrontierSWE	Epoch AI Benchmarking Hub	–	25.9%
Terminal-Bench 4.0 (AA)	Artificial Analysis	6.6%	12.6%
Terminal-Bench 2.1 (AA)	Artificial Analysis	78.7%	85.0%
Startrise Frontend	Startrise	–	79.8%†
FrontierCode 1.1 (AA)	Artificial Analysis	–	43.6%
Agents & tools
APEX-Agents	Mercor	27.5%†	47.8%†
Arena Agent	Arena (LMArena)	–	1.6
LiveBench Long Agentic	LiveBench	–	42.8%†
LiveBench Browser	LiveBench	–	16.0%†
LiveBench Agentic Coding	LiveBench	49.0%	62.2%†
Terminal-Bench Hard	Artificial Analysis	40.9%	–
τ²-Bench Telecom (AA)	Artificial Analysis	95.3%	–
Terminal-Bench 2.1 (Vals)	Vals AI	74.2%	80.9%†
MCP Atlas	Scale AI SEAL	83.6%	82.3%
HiL-Bench	Scale AI SEAL	27.7%†	42.7%†
τ³-bench Banking	τ²-bench	–	37.1%
τ³-Banking (AA)	Artificial Analysis	32.2%	46.0%
ITBench SRE (AA)	Artificial Analysis	40.4%	47.7%
Analyst Agent (AA)	Artificial Analysis	45.0%	38.8%
APEX-Agents (AA)	Artificial Analysis	47.0%	41.3%
CyberBench	Vals AI	–	75.2%
Finance Agent v2	Vals AI	57.9%	53.1%
Finance Agent v2.1	Vals AI	–	60.0%
Harvey's Legal Agent Benchmark v1.1	Vals AI	–	12.9%
Harvey's Legal Agent Benchmark v1	Vals AI	2.5%	12.9%
Legal Research Bench	Vals AI	30.8%	46.1%
SkillsBench	Vals AI	52.7%	–
Tax Agent Bench	Vals AI	21.6%	33.4%
Terminal-Bench 2.0 (Vals)	Vals AI	67.4%	–
Terminal-Bench 4.0 (Vals)	Vals AI	6.1%	17.2%
Terminal-Bench Science	Vals AI	5.7%	1.4%
Time Horizon Index: KSP	Vals AI	–	10.5%
Web Search Index	Vals AI	37.0%	–
PostTrainBench	Epoch AI Benchmarking Hub	–	32.0%†
GBAEval	Epoch AI Benchmarking Hub	6.7%†	48.4%†
Vending-Bench 2	Andon Labs	5396.4†	5165.0†
AutomationBench	Artificial Analysis	42.1%	58.3%
GDP.pdf (AA)	Artificial Analysis	19.8%	22.0%
Harvey LAB-AA v1.1	Harvey	–	5.3%
EnterpriseOps-Gym	Artificial Analysis	50.1%	45.3%
AA-Briefcase v1.1	Artificial Analysis	874	1499
Terminal-Bench-Science 0.1 (AA)	Artificial Analysis	–	8.1%
CWE-Bench (AA)	Artificial Analysis	–	53.3%
DeepSec-Bench (AA)	Artificial Analysis	–	12.4%
CyberGym end-to-end (AA)	Artificial Analysis	–	58.0%
FACTS Search	Google	64.3%†	–
Maths
FrontierMath Tiers 1–3	Epoch AI Benchmarking Hub	62.8%	72.2%
FrontierMath Tier 4	Epoch AI Benchmarking Hub	26.8%	39.0%
OTIS Mock AIME	Epoch AI Benchmarking Hub	95.6%	97.2%
ProofBench	Vals AI	31.0%	87.0%†
LiveBench Mathematics	LiveBench	88.2%	84.4%†
AIME 2026	MathArena	95.0%†	96.7%†
HMMT February 2026	MathArena	95.5%†	97.0%†
MathArena Apex	MathArena	32.3%†	65.6%†
Arena Maths	Arena (LMArena)	1497	1499
Knowledge
SimpleQA Verified	Epoch AI Benchmarking Hub	66.2%	50.6%
LiveBench 2.0 Data Analysis	LiveBench	–	78.7%†
LiveBench Research	LiveBench	–	79.0%†
LiveBench 1.0 Data Analysis	LiveBench	64.9%	78.7%†
AA-Omniscience	Artificial Analysis	21.2	19.7
MMLU-Pro	Vals AI	89.5%	88.0%
LegalBench	Vals AI	83.6%	86.2%
CorpFin	Vals AI	64.7%	71.6%†
TaxEval	Vals AI	74.4%	75.7%†
BioMysteryBench	Vals AI	–	72.6%
Excel Modeling Benchmark	Vals AI	63.5%	66.7%
Excel Modeling Benchmark v1.1	Vals AI	–	71.3%
MedCode	Vals AI	55.8%	49.4%
MedScribe	Vals AI	76.6%	88.0%
Public Benefits Bench v1.1	Vals AI	59.5%	68.2%
DrugDiscoveryBench	Scale AI SEAL	50.0%	–
EBR-bench	Epoch AI Benchmarking Hub	4.8%	–
GDP.pdf	Epoch AI Benchmarking Hub	14.0%†	19.0%
AA-Omniscience: accuracy	Artificial Analysis	51.4%	47.6%
AA-Omniscience: non-hallucination	Artificial Analysis	37.8%	46.8%
Instruction following
LiveBench Language	LiveBench	84.6%	85.5%†
LiveBench Instruction Following	LiveBench	75.6%	71.4%†
IFBench	Artificial Analysis	76.3%	–
Arena Instruction Following	Arena (LMArena)	1465	1489
Human preference
Arena Text	Arena (LMArena)	1477	1487
EQ-Bench 4	EQ-Bench	1081†	1318†
Arena Creative Writing	Arena (LMArena)	1464	1462
Arena Multi-turn	Arena (LMArena)	1480	1502
Arena Longer Queries	Arena (LMArena)	1482	1503
Multimodal
Arena Vision	Arena (LMArena)	1308	–
LiveBench Design	LiveBench	–	77.2%†
MMMU-Pro	Artificial Analysis	84.3%	80.5%
MMMU-Pro (Vals)	Vals AI	88.3%	88.2%†
MortgageTax	Vals AI	68.1%	66.3%†
SAGE	Vals AI	49.9%	52.8%
Blueprint-Bench 2	Epoch AI Benchmarking Hub	33.6%†	29.5%†
AA-3D	Artificial Analysis	–	982
Long context
AA-LCR	Artificial Analysis	73.3%	88.7%
Arena Document	Arena (LMArena)	1453	–
MLCR	Artificial Analysis	10.0%	21.7%
Composite
Epoch Capabilities Index	Epoch AI Benchmarking Hub	154.9†	157.9†
LiveBench 2.0	LiveBench	–	63.4%†
LiveBench 1.0	LiveBench	74.6%	79.2%†
AA Intelligence Index v4.3.2	Artificial Analysis	32.6	43.6
Vals Index	Vals AI	–	53.0
Vals Finance Index	Vals AI	–	58.0%
Vals RSI Index	Vals AI	–	20.9%
Vals Multimodal Index	Vals AI	62.9%	73.4%†
Vals Router Index	Vals AI	–	53.0%†
AA Cyber Index	Artificial Analysis	–	41.3
Safety & honesty
FORTRESS	Scale AI SEAL	–	26.6%
Metadata
AA Openness Index	Artificial Analysis	–	38.9

Data as of 2026-10-25. Best configuration of each model; every score links to its source on the model pages.

What changed

23 Oct 2026 Gemini 3.5 Flash (high): 1 new result (Harvey's Legal Agent Benchmark v1).
23 Oct 2026 Kimi K3 (max): 7 new results (Code Migration v1.1, Excel Modeling Benchmark v1.1, Finance Agent v2.1…).
21 Oct 2026 Gemini 3.5 Flash (high): 1 score revision.
21 Oct 2026 Kimi K3 (max): 1 new result (Terminal-Bench-Science 0.1 (AA)); 1 score revision.
20 Oct 2026 Kimi K3 (max): 3 score revisions.
19 Oct 2026 Gemini 3.5 Flash (high): BenchLeader Index 62.5 → 63.2.
19 Oct 2026 Kimi K3 (max): 1 score revision.
18 Oct 2026 Gemini 3.5 Flash (high): 3 new results (GDPval-AA v2.1, AA-Omniscience: accuracy, AA-Omniscience: non-hallucination).
18 Oct 2026 Kimi K3 (max): 1 new result (τ³-bench Banking).
16 Oct 2026 Gemini 3.5 Flash (high): 1 score revision.

Rankings: Best LLM overall · Best LLM for agents · Best cheap LLMs