GPT-3.5 Turbo vs GPT-5

GPT-3.5 Turbo vs GPT-5: GPT-3.5 Turbo is cheaper by 85% on average. GPT-3.5 Turbo from OpenAI (16,385-token context, tool calls) vs. GPT-5 from OpenAI (272,000-token context, reasoning, tool calls). Use Agent Command Center to A/B both in shadow mode and pick the winner per workload.

Side-by-side cost

Live workload comparison

Same workload run through both models. The cheaper one is highlighted.

3,000
0272,000
400
0128,000
5,000
01,000,000
OpenAI
$320/mo
Input $0.500/M · Output $1.50/M
OpenAI
$1,179/mo
Input $1.25/M · Output $10.00/M
At this workload, GPT-3.5 Turbo is 73% cheaper than GPT-5 — a savings of $860/month ($10,318/year).
Production recipe — Agent Command Center
strategy: cost-optimized
primary:
  model: gpt-3-5-turbo
  provider: openai
fallback:
  model: gpt-5
  provider: openai
shadow: { sample_rate: 0.05 }   # mirror 5% of traffic to compare quality live
GPT-3.5 Turbo GPT-5
Input price $0.500/M $1.25/M
Output price $1.50/M $10.00/M
Context window 16,385 272,000
Max output 4,096 128,000
Function calling
Vision
Audio input
Reasoning
Prompt caching
Structured output
Pricing verified May 12, 2026 May 12, 2026
Cheaper option
~85% cheaper than GPT-5
Larger context
272,000 tokens
More capabilities
5 of 6 capability flags advertised

Benchmark comparison

Side-by-side public benchmark scores. Greener bar = winner.

Chatbot Arena ELOgeneral
GPT-3.5 Turbo
GPT-5
1,450
MATH-500math
GPT-3.5 Turbo
GPT-5
99.6%
AIME 2024math
GPT-3.5 Turbo
GPT-5
98.4%
BFCL v3agent
GPT-3.5 Turbo
GPT-5
96.3%
HumanEvalcode
GPT-3.5 Turbo
GPT-5
96.0%
IFEvalgeneral
GPT-3.5 Turbo
GPT-5
95.6%
AIME 2025math
GPT-3.5 Turbo
GPT-5
94.6%
LiveCodeBenchcode
GPT-3.5 Turbo
GPT-5
90.0%
MMLU-Proreasoning
GPT-3.5 Turbo
GPT-5
89.4%
Aider Polyglotcode
GPT-3.5 Turbo
GPT-5
88.0%
GPQA Diamondreasoning
GPT-3.5 Turbo
GPT-5
87.3%
MMMUmultimodal
GPT-3.5 Turbo
GPT-5
84.2%
SWE-bench Verifiedagent
GPT-3.5 Turbo
GPT-5
74.9%
Humanity's Last Examreasoning
GPT-3.5 Turbo
GPT-5
42.0%
ARC-AGI-2reasoning
GPT-3.5 Turbo
GPT-5
17.6%