Compare · ModelsLive · 2 picked · head to head
Gemini 2.5 Pro vs GPT-6 Luna
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-6 Luna wins on 11/12 benchmarks
GPT-6 Luna wins 11 of 12 shared benchmarks. Leads in speed · agentic · reasoning.
Category leads
speed·GPT-6 Lunaagentic·GPT-6 Lunareasoning·GPT-6 Lunaknowledge·GPT-6 Lunageneral·GPT-6 Lunamath·GPT-6 Luna
Hype vs Reality
Attention vs performance
Gemini 2.5 Pro
#124 by perf·no signal
GPT-6 Luna
#103 by perf·#5 by attention
Best value
GPT-6 Luna
19.6x better value than Gemini 2.5 Pro
Gemini 2.5 Pro
9.0 pts/$
$5.63/M
GPT-6 Luna
176.7 pts/$
$0.30/M
Vendor risk
Who is behind the model
Google DeepMind
$4.27T·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
12 benchmarks · 2 models
Gemini 2.5 ProGPT-6 Luna
Artificial Analysis · Quality Index
GPT-6 Luna leads by +11.1
Gemini 2.5 Pro
27.0
GPT-6 Luna
38.1
APEX-Agents
GPT-6 Luna leads by +37.7
APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments.
Gemini 2.5 Pro
6.6
GPT-6 Luna
44.3
ARC-AGI
GPT-6 Luna leads by +45.7
ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization.
Gemini 2.5 Pro
41.0
GPT-6 Luna
86.7
ARC-AGI-2
GPT-6 Luna leads by +54.5
ARC-AGI-2 · the second iteration of the Abstraction and Reasoning Corpus, testing novel pattern recognition and abstract reasoning without prior training data.
Gemini 2.5 Pro
4.9
GPT-6 Luna
59.3
Chess Puzzles
GPT-6 Luna leads by +16.8
Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities.
Gemini 2.5 Pro
15.8
GPT-6 Luna
32.7
Dtbench
GPT-6 Luna leads by +12.9
Gemini 2.5 Pro
70.7
GPT-6 Luna
83.5
FrontierMath-Tier-4-v2-Private
GPT-6 Luna leads by +56.1
Gemini 2.5 Pro
0.0
GPT-6 Luna
56.1
FrontierMath-Tiers-1-3-v2-Private
GPT-6 Luna leads by +54.4
Gemini 2.5 Pro
24.6
GPT-6 Luna
79.0
GPQA diamond
GPT-6 Luna leads by +6.9
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
Gemini 2.5 Pro
80.4
GPT-6 Luna
87.3
Lmca
GPT-6 Luna leads by +11.4
Gemini 2.5 Pro
40.9
GPT-6 Luna
52.3
OTIS Mock AIME 2024-2025
GPT-6 Luna leads by +14.2
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
Gemini 2.5 Pro
84.7
GPT-6 Luna
98.9
SimpleQA Verified
Gemini 2.5 Pro leads by +14.6
SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information.
Gemini 2.5 Pro
56.0
GPT-6 Luna
41.4
Full benchmark table
| Benchmark | Gemini 2.5 Pro | GPT-6 Luna |
|---|---|---|
Artificial Analysis · Quality Index | 27.0 | 38.1 |
APEX-Agents APEX-Agents · evaluates AI agents on complex, multi-step tasks requiring planning, tool use, and autonomous decision-making in realistic environments. | 6.6 | 44.3 |
ARC-AGI ARC-AGI · the original Abstraction and Reasoning Corpus, testing whether AI can solve novel visual pattern recognition tasks without memorization. | 41.0 | 86.7 |
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. | 4.9 | 59.3 |
Chess Puzzles Chess Puzzles · tests strategic and tactical reasoning by having models solve chess puzzle positions, evaluating lookahead and pattern recognition abilities. | 15.8 | 32.7 |
Dtbench | 70.7 | 83.5 |
FrontierMath-Tier-4-v2-Private | 0.0 | 56.1 |
FrontierMath-Tiers-1-3-v2-Private | 24.6 | 79.0 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 80.4 | 87.3 |
Lmca | 40.9 | 52.3 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 84.7 | 98.9 |
SimpleQA Verified SimpleQA Verified · short factual questions with verified answers, measuring factual accuracy and the tendency to hallucinate or provide incorrect information. | 56.0 | 41.4 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $1.25 | $10.00 | 1.0M tokens (~524 books) | $34.38 | |
| $0.10 | $0.50 | 1.1M tokens (~525 books) | $2.00 |