VerdictPal · editorial desk · 2026VerdictPal

Compare · Two models, head-to-head

Pick two models. See the diff.

Composite scores, benchmark ranks, capability grids, and pricing per million tokens. All sourced from the same data that powers the model atlas.

Cards Models

Gemini 3.5 Flash vs Claude Sonnet 4.6

Model A

Model B

MODEL A

Gemini 3.5 Flash

Google · Gemini

Composite 67Intel 4927 rows11 sources

Google's fastest frontier-tier model. Leads the MCP Atlas tool-use benchmark (83.6%) and pairs a 1M-token context with sub-200ms first-token latency. Terminal-Bench 2.0 76.2%, ARC-AGI-2 72.1%. $0.30 / $1.20 per 1M tokens.

MODEL B

Claude Sonnet 4.6

Anthropic · Claude Sonnet

Composite 66Intel 6024 rows11 sources

Anthropic's February 2026 mid-tier model. Vals Index 60.30% (#3 on the public leaderboard — beats GPT 5.4 Mini and Gemini 3 Flash). SWE-bench Verified 79.6%, ARC-AGI-2 58.3%. 1M context, $3 / $15 per 1M tokens — 5× cheaper than Opus 4.8 on input.

Gemini 3.5 Flash3stats won

Tied1stats

Claude Sonnet 4.63stats won

Spec showdown

Capacity, price, freshness.

Gemini 3.5 FlashvsClaude Sonnet 4.6

Composite scores are within 5 points — 67 for Gemini 3.5 Flash vs 66 for Claude Sonnet 4.6. The choice comes down to capability profile and price. Claude Sonnet 4.6 has the higher Vals/AA intelligence score (60 vs 49). Gemini 3.5 Flash is significantly cheaper at $0.30/1M input vs $3.0/1M input.

Composite score
Gemini 3.5 Flash67Winner on this rowClaude Sonnet 4.666
Intelligence (Vals/AA)
Gemini 3.5 Flash49Claude Sonnet 4.660Winner on this row
Context window
Gemini 3.5 Flash1MClaude Sonnet 4.61M
Max output
Gemini 3.5 Flash66KClaude Sonnet 4.6128KWinner on this row
Input price / 1M (lower wins)
Gemini 3.5 Flash$0.30Winner on this rowClaude Sonnet 4.6$3.0
Output speed (AA default API)
Gemini 3.5 Flash287 t/sWinner on this rowClaude Sonnet 4.660 t/s
Time to first token (AA default API)
Gemini 3.5 Flash12.11sClaude Sonnet 4.61.18sWinner on this row

Composite score

Intelligence (Vals/AA)

Context window

Max output

66K

128K

Input price / 1M (lower wins)

$0.30

$3.0

Output speed (AA default API)

287 t/s

60 t/s

Time to first token (AA default API)

12.11s

1.18s

Capability grid

What each model can do.

Modalities

Gemini 3.5 FlashClaude Sonnet 4.6

Text●●

Image●—

Audio●—

Video●—

Code●●

Tool use●●

Reasoning●●

Access

Gemini 3.5 FlashClaude Sonnet 4.6

API●●

Consumer app●●

Open weights——

Managed cloud●●

On-device——

Family radar

Top score per benchmark family.

Knowledge

Reasoning

Coding

Agentic

Multimodal

Long context

Tool use

Human preference

Benchmark ranks

Every shared benchmark, side by side.

Knowledge

MMLU#12/22 · #4/22

88.4%

90.6%

Reasoning

ARC-AGI#7/14 · #8/14

72.1%

58.3%

Artificial Analysis Intelligence Index#16/67 · #43/67

50.2%

35.9%

GPQA Diamond#59/69 · #54/69

74.1%

77.2%

Humanity's Last Exam#11/65 · #54/65

41%

13.2%

IFBench#11/57 · #54/57

76.33%

41.16%

LiveBench#8/21 · #5/21

64.2%

67.2%

Coding

HumanEval#6/15 · #7/15

94.6% pass@1

94.1% pass@1

SciCode#18/65 · #37/65

53.1%

46.9%

SWE-bench Verified#20/23 · #13/23

77.2%

79.6%

Agentic

Terminal-Bench#16/69 · #24/69

76.2%

68.4%

τ³-Bench Banking#35/64 · #12/64

25.36%

79.53%

Multimodal

MMMU#4/8 · #5/8

68.9%

66.2%

Long context

AA-LCR#25/64 · #51/64

69.33%

57.67%

LongBench v2#7/19 · #8/19

53.8%

52.6%

Tool use

BFCL#6/20 · #1/20

82.3%

88.7%

MCP Atlas#1/10 · #4/10

83.6%

75.4%

Human preference

LMArena (Chatbot Arena)#3/13 · #10/13

1455

1396

Performance

AA output speed#6/69 · #48/69

287 t/s

60 t/s

AA time to first token#19/69 · #39/69

12.11s

1.18s

Dossier facts

Dates, sources, fineprint.

Gemini 3.5 FlashClaude Sonnet 4.6

ProviderGoogleAnthropic

FamilyGeminiClaude Sonnet

Release date2026-05-192026-02-17

Knowledge cutoff2026-03-012025-12-01

Context window1M1M

Max output66K128K

Input price /1M$0.30$3.0

Output price /1M$1.2$15

Statusflagshipflagship

Composite score6766

Intelligence score4960

Intel sourcevalsvals

Benchmark rows2724

Source count1111

Freshnessfreshfresh

Pricing checked2026-06-012026-06-01

Ecosystem

Which tools wrap each model.

Gemini 3.5 Flash

Claude Sonnet 4.6

Verdict

Which model to actually pick.

Open Gemini 3.5 Flash's dossier Open Claude Sonnet 4.6's dossier