Compare · ModelsLive · 2 picked · head to head
Gemini 3.7 Flash vs Kimi K3
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Gemini 3.7 Flash wins on 9/18 benchmarks
Gemini 3.7 Flash wins 9 of 18 shared benchmarks. Leads in agentic · reasoning · knowledge.
Category leads
agentic·Gemini 3.7 Flashreasoning·Gemini 3.7 Flashknowledge·Gemini 3.7 Flashcoding·Kimi K3general·Kimi K3math·Kimi K3
Hype vs Reality
Attention vs performance
Gemini 3.7 Flash
#59 by perf·#10 by attention
Kimi K3
#67 by perf·#15 by attention
Best value
Gemini 3.7 Flash
3.6x better value than Kimi K3
Gemini 3.7 Flash
25.9 pts/$
$2.25/M
Kimi K3
7.1 pts/$
$8.10/M
Vendor risk
Who is behind the model
Google DeepMind
$3.53T·Tier 1
Moonshot AI
$50.0B·Tier 1
Head to head
18 benchmarks · 2 models
Gemini 3.7 FlashKimi K3
APEX-Agents
Gemini 3.7 Flash leads by +20.0
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Gemini 3.7 Flash
67.8
Kimi K3
47.8
ARC-AGI
Gemini 3.7 Flash leads by +1.0
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Gemini 3.7 Flash
95.5
Kimi K3
94.5
ARC-AGI-2
Gemini 3.7 Flash leads by +24.2
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Gemini 3.7 Flash
84.6
Kimi K3
60.4
Chess Puzzles
Gemini 3.7 Flash leads by +30.5
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Gemini 3.7 Flash
64.2
Kimi K3
33.7
Deepswe
Kimi K3 leads by +3.0
Gemini 3.7 Flash
65.5
Kimi K3
68.5
Dtbench
Gemini 3.7 Flash leads by +9.3
Gemini 3.7 Flash
94.7
Kimi K3
85.3
Frontiercode
Kimi K3 leads by +0.6
Gemini 3.7 Flash
43.6
Kimi K3
44.2
FrontierMath-Tier-4-v2-Private
Kimi K3 leads by +2.4
Gemini 3.7 Flash
36.6
Kimi K3
39.0
FrontierMath-Tiers-1-3-v2-Private
Kimi K3 leads by +0.6
Gemini 3.7 Flash
75.3
Kimi K3
76.0
Frontierswe
Kimi K3 leads by +5.6
Gemini 3.7 Flash
20.3
Kimi K3
25.9
Furniture Assembly
Kimi K3 leads by +6.0
Gemini 3.7 Flash
0.0
Kimi K3
6.0
GPQA diamond
Gemini 3.7 Flash leads by +2.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemini 3.7 Flash
97.0
Kimi K3
94.6
Lmca
Kimi K3 leads by +2.8
Gemini 3.7 Flash
59.3
Kimi K3
62.0
Mirrorcode
Kimi K3 leads by +13.3
Gemini 3.7 Flash
41.1
Kimi K3
54.4
Mystery Game Puzzles
Gemini 3.7 Flash leads by +12.1
Gemini 3.7 Flash
30.6
Kimi K3
18.5
OTIS Mock AIME 2024-2025
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemini 3.7 Flash
97.2
Kimi K3
97.2
Proofbench
Kimi K3 leads by +29.0
Gemini 3.7 Flash
58.0
Kimi K3
87.0
SimpleQA Verified
Gemini 3.7 Flash leads by +18.6
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Gemini 3.7 Flash
69.2
Kimi K3
50.6
Full benchmark table
| Benchmark | Gemini 3.7 Flash | Kimi K3 |
|---|---|---|
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 67.8 | 47.8 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 95.5 | 94.5 |
ARC-AGI-2 ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data. | 84.6 | 60.4 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 64.2 | 33.7 |
Deepswe | 65.5 | 68.5 |
Dtbench | 94.7 | 85.3 |
Frontiercode | 43.6 | 44.2 |
FrontierMath-Tier-4-v2-Private | 36.6 | 39.0 |
FrontierMath-Tiers-1-3-v2-Private | 75.3 | 76.0 |
Frontierswe | 20.3 | 25.9 |
Furniture Assembly | 0.0 | 6.0 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 97.0 | 94.6 |
Lmca | 59.3 | 62.0 |
Mirrorcode | 41.1 | 54.4 |
Mystery Game Puzzles | 30.6 | 18.5 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 97.2 | 97.2 |
Proofbench | 58.0 | 87.0 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 69.2 | 50.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.75 | $3.75 | 1.0M tokens (~524 books) | $15.00 | |
| $0.19 | $16.00 | 1.0M tokens (~524 books) | $41.44 |