Which is better, Gemini 3.5 Flash or Kimi K2.6?

Across 46 shared benchmarks tracked on Vector Wire, Gemini 3.5 Flash leads on 36 and Kimi K2.6 leads on 9 — Gemini 3.5 Flash leads on more. Strengths differ by axis; the table below shows every head-to-head number with provenance.

SEEDRefresh Models —Deals —Regulatory —Sources —

Gemini 3.5 Flash vs Kimi K2.6

GooglevsMoonshot46 shared benchmarks36–9 head-to-head

Benchmark	Gemini 3.5 Flash	Kimi K2.6
AA Agentic Index	70.4	58.7
AA Intelligence	52	45.1
AA-LCR	81	76.7
AA-Omniscience	21.2	6.4
aime_2026	95	96.4
apexAgents	47.1	28.5
arena_elo	1477	1461
arena_text_factuality	1466	1455
Artificial Analysis Coding Index	70.1	61.8
charxiv_rq	84.2	86.7
coding_arena_elo	1490	1509
critpt	13.1	8
Finance Agent v2	57.9	44.9
frontiermath_tier_4	14.6	14.6
gdpval	57.8	41.4
GDPval-AA (Elo)	1656	1482
GDPval-AA v2	1357	1190
GPQA Diamond	92.2	91.1
HiL-Bench	27.7	18.7
HLE	42.7	37.5
hmmt_feb_2026	95.5	94.7
IFBench	76.3	76
itbenchSre	40.3	31.2
livebench	75	72.2
livebench_agentic_coding	49	46.9
livebench_coding	78.2	78.6
livebench_data_analysis	64.9	65.1
livebench_instruction_following	75.6	64.4
livebench_language	84.6	75.1
livebench_math	88.2	84.3
livebench_reasoning	82	79.4
MCP Atlas	83.6	68.1
MMMU-Pro	84.3	80.1
OmniScience Accuracy	51.4	32.8
OmniScience Non-Hallucination	38.2	60.7
OSWorld-Verified	78.4	73.1
scicode	53.1	53.5
simpleqa_verified	68.4	38.7
SWE-bench Pro	55.1	58.6
TauBench V3 - Banking	32.2	23.3
Terminal-Bench 2.0	76.2	66.7
Terminal-Bench 2.1	78.7	65.9
Terminal-Bench Hard	46.2	43.9
toolathlon	56.5	50
τ²-Bench Telecom (AA run)	95.6	95.9
τ³-Bench	25.4	20.6

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.