Compare · ModelsLive · 2 picked · head to head
GPT-4.5 vs Gemma 3 27B (free)
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-4.5 wins on 5/6 benchmarks
GPT-4.5 wins 5 of 6 shared benchmarks. Leads in coding · knowledge · math.
Category leads
coding·GPT-4.5knowledge·GPT-4.5math·GPT-4.5
Hype vs Reality
Attention vs performance
GPT-4.5
#166 by perf·no signal
Gemma 3 27B (free)
#131 by perf·no signal
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
Google DeepMind
$4.00T·Tier 1
Head to head
6 benchmarks · 2 models
GPT-4.5Gemma 3 27B (free)
Aider polyglot
GPT-4.5 leads by +40.0
Aider Polyglot · measures how well AI models can edit code across multiple programming languages using the Aider coding assistant framework.
GPT-4.5
44.9
Gemma 3 27B (free)
4.9
Fiction.LiveBench
GPT-4.5 leads by +30.6
Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination.
GPT-4.5
63.9
Gemma 3 27B (free)
33.3
GPQA diamond
GPT-4.5 leads by +26.4
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GPT-4.5
58.3
Gemma 3 27B (free)
31.8
Lech Mazur Writing
Gemma 3 27B (free) leads by +4.3
Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication.
GPT-4.5
75.6
Gemma 3 27B (free)
79.9
MATH level 5
GPT-4.5 leads by +4.6
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
GPT-4.5
78.6
Gemma 3 27B (free)
74.0
OTIS Mock AIME 2024-2025
GPT-4.5 leads by +18.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-4.5
37.7
Gemma 3 27B (free)
19.6
Full benchmark table
| Benchmark | GPT-4.5 | Gemma 3 27B (free) |
|---|---|---|
Aider polyglot Aider Polyglot · measures how well AI models can edit code across multiple programming languages using the Aider coding assistant framework. | 44.9 | 4.9 |
Fiction.LiveBench Fiction.LiveBench · a continuously updated benchmark using recently published fiction to test reading comprehension and reasoning, preventing data contamination. | 63.9 | 33.3 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 58.3 | 31.8 |
Lech Mazur Writing Lech Mazur Writing · evaluates creative writing ability, assessing prose quality, narrative coherence, and stylistic sophistication. | 75.6 | 79.9 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | 78.6 | 74.0 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 37.7 | 19.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| — | — | — | — | |
| $0.00 | $0.00 | 131K tokens (~66 books) | — |