Compare/GPT-5.4 (xhigh) vs Gemini 3 Pro Preview (high)

GPT-5.4 (xhigh)vsGemini 3 Pro Preview (high)

Side-by-side comparison of pricing, 12 benchmarks, and generation speed.

OpenAI

GPT-5.4 (xhigh)

Input
$2.5/M
Output
$15/M
Speed
84 tok/s
TTFT
147.97s
Google

Gemini 3 Pro Preview (high)

Input
$2/M
Output
$12/M
Speed
117 tok/s
TTFT
39.61s

Winner by Category

Cheaper
Gemini 3 Pro Preview (high)
Faster (tok/s)
Gemini 3 Pro Preview (high)
Lower Latency
Gemini 3 Pro Preview (high)
Benchmarks (7-4)
GPT-5.4 (xhigh)

Pricing Comparison

MetricGPT-5.4 (xhigh)Gemini 3 Pro Preview (high)
Input ($/M tokens)$2.5$2
Output ($/M tokens)$15$12
Cost for 1M input + 100K output tokens:
GPT-5.4 (xhigh)$4.00
Gemini 3 Pro Preview (high)$3.20

Speed Comparison

Output Speed (tokens/s) — higher is better
GPT-5.4 (xhigh)
84 tok/s
Gemini 3 Pro Preview (high)
117 tok/s
Time to First Token (seconds) — lower is better
GPT-5.4 (xhigh)
147.97s
Gemini 3 Pro Preview (high)
39.61s

Benchmark Comparison

Data from Artificial Analysis API — 12 benchmarks

Intelligence Index
57.248.4
Coding Index
57.346.5
Math Index
95.7
GPQA Diamond
92.0%90.8%
MMLU-Pro
89.8%
LiveCodeBench
91.7%
AIME 2025
95.7%
MATH-500
Humanity's Last Exam
41.6%37.2%
SciCode
56.6%56.1%
IFBench
73.9%70.4%
TerminalBench
57.6%41.7%
GPT-5.4 (xhigh)7 wins
4 winsGemini 3 Pro Preview (high)

Frequently Asked Questions

Which is cheaper, GPT-5.4 (xhigh) or Gemini 3 Pro Preview (high)?

Gemini 3 Pro Preview (high) is cheaper overall. Its blended price (3:1 input/output ratio) is $4.50/M tokens vs $5.63/M for GPT-5.4 (xhigh).

Which model performs better on benchmarks?

GPT-5.4 (xhigh) wins 7 out of 12 benchmarks compared to 4 for Gemini 3 Pro Preview (high). See the detailed benchmark chart above for per-category results.

Which is faster for real-time applications?

Gemini 3 Pro Preview (high) generates tokens faster at 117 tok/s vs 84 tok/s. However, Gemini 3 Pro Preview (high) has lower time-to-first-token (39.61s vs 147.97s).

When should I use GPT-5.4 (xhigh) vs Gemini 3 Pro Preview (high)?

Choose based on your priorities: Gemini 3 Pro Preview (high) for lower cost, GPT-5.4 (xhigh) for stronger benchmark performance, and Gemini 3 Pro Preview (high) for faster generation. For latency-sensitive apps, check the TTFT comparison above.