Compare · ModelsLive · 2 picked · head to head
GPT-3.5 Turbo (older v0613) vs GPT-4 Turbo (batch)
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
GPT-4 Turbo (batch) wins on 11/13 benchmarks
GPT-4 Turbo (batch) wins 11 of 13 shared benchmarks. Leads in reasoning · general · math.
Category leads
reasoning·GPT-4 Turbo (batch)general·GPT-4 Turbo (batch)math·GPT-4 Turbo (batch)knowledge·GPT-4 Turbo (batch)coding·GPT-4 Turbo (batch)
Hype vs Reality
Attention vs performance
GPT-3.5 Turbo (older v0613)
#250 by perf·no signal
GPT-4 Turbo (batch)
#188 by perf·no signal
Best value
GPT-3.5 Turbo (older v0613)
5.2x better value than GPT-4 Turbo (batch)
GPT-3.5 Turbo (older v0613)
21.9 pts/$
$1.50/M
GPT-4 Turbo (batch)
4.2 pts/$
$10.00/M
Vendor risk
Who is behind the model
OpenAI
$840.0B·Tier 1
OpenAI
$840.0B·Tier 1
Head to head
13 benchmarks · 2 models
GPT-3.5 Turbo (older v0613)GPT-4 Turbo (batch)
BBH
GPT-4 Turbo (batch) leads by +18.0
BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans.
GPT-3.5 Turbo (older v0613)
48.8
GPT-4 Turbo (batch)
66.8
Dtbench
GPT-4 Turbo (batch) leads by +23.6
GPT-3.5 Turbo (older v0613)
14.2
GPT-4 Turbo (batch)
37.8
FrontierMath-Tiers-1-3-v2-Private
GPT-4 Turbo (batch) leads by +0.7
GPT-3.5 Turbo (older v0613)
0.0
GPT-4 Turbo (batch)
0.7
GPQA diamond
GPT-4 Turbo (batch) leads by +4.8
Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs.
GPT-3.5 Turbo (older v0613)
3.0
GPT-4 Turbo (batch)
7.8
GSM8K
GPT-4 Turbo (batch) leads by +32.2
Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve.
GPT-3.5 Turbo (older v0613)
57.8
GPT-4 Turbo (batch)
90.0
Lmca
GPT-4 Turbo (batch) leads by +8.7
GPT-3.5 Turbo (older v0613)
11.3
GPT-4 Turbo (batch)
20.1
MATH level 5
GPT-4 Turbo (batch) leads by +11.3
MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics.
GPT-3.5 Turbo (older v0613)
11.6
GPT-4 Turbo (batch)
23.0
MMLU
GPT-4 Turbo (batch) leads by +20.1
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
GPT-3.5 Turbo (older v0613)
56.4
GPT-4 Turbo (batch)
76.5
Mystery Game Puzzles
GPT-4 Turbo (batch) leads by +3.1
GPT-3.5 Turbo (older v0613)
0.0
GPT-4 Turbo (batch)
3.1
OTIS Mock AIME 2024-2025
GPT-3.5 Turbo (older v0613) leads by +1.1
OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills.
GPT-3.5 Turbo (older v0613)
2.1
GPT-4 Turbo (batch)
1.0
TriviaQA
GPT-3.5 Turbo (older v0613) leads by +1.0
TriviaQA · reading comprehension benchmark with trivia questions, requiring models to find and reason over evidence from provided documents.
GPT-3.5 Turbo (older v0613)
85.8
GPT-4 Turbo (batch)
84.8
WeirdML
GPT-4 Turbo (batch) leads by +8.9
WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns.
GPT-3.5 Turbo (older v0613)
3.5
GPT-4 Turbo (batch)
12.4
Winogrande
GPT-4 Turbo (batch) leads by +11.8
WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs.
GPT-3.5 Turbo (older v0613)
63.2
GPT-4 Turbo (batch)
75.0
Full benchmark table
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo (batch) |
|---|---|---|
BBH BIG-Bench Hard · a curated subset of 23 challenging tasks from BIG-Bench where language models previously failed to outperform average humans. | 48.8 | 66.8 |
Dtbench | 14.2 | 37.8 |
FrontierMath-Tiers-1-3-v2-Private | 0.0 | 0.7 |
GPQA diamond Graduate-Level Google-Proof QA (Diamond set) · expert-crafted questions in physics, biology, and chemistry that are difficult even for domain PhDs. | 3.0 | 7.8 |
GSM8K Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve. | 57.8 | 90.0 |
Lmca | 11.3 | 20.1 |
MATH level 5 MATH Level 5 · the hardest tier of the MATH benchmark, featuring competition-level problems from AMC, AIME, and Olympiad-style mathematics. | 11.6 | 23.0 |
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge. | 56.4 | 76.5 |
Mystery Game Puzzles | 0.0 | 3.1 |
OTIS Mock AIME 2024-2025 OTIS Mock AIME 2024-2025 · simulated American Invitational Mathematics Examination problems testing advanced problem-solving skills. | 2.1 | 1.0 |
TriviaQA TriviaQA · reading comprehension benchmark with trivia questions, requiring models to find and reason over evidence from provided documents. | 85.8 | 84.8 |
WeirdML WeirdML · tests models on unusual and adversarial machine learning tasks that require creative problem-solving beyond standard patterns. | 3.5 | 12.4 |
Winogrande WinoGrande · large-scale commonsense reasoning benchmark where models must resolve ambiguous pronouns in carefully constructed sentence pairs. | 63.2 | 75.0 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $1.00 | $2.00 | 4K tokens (~2 books) | $12.50 | |
| $5.00 | $15.00 | 128K tokens (~64 books) | $75.00 |