ModelsCompareBest forBenchmarksStatusPricingAPI
Compare / head-to-head

GPT-5.4 minivsGPT-5.5

GPT-5.5 leads 8 of 9 shared benchmarks. GPT-5.4 mini is 6.7x cheaper per token. GPT-5.5 has the larger context window (1M tokens).

Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.

Shared benchmarks
1 – 8
GPT-5.5 leads
Cheaper per token
GPT-5.4 mini
6.7x cheaper, input + output
Larger context
GPT-5.5
1M tokens
Provider uptime (30d)
94.2446%
OpenAI
GPT-5.4 mini
OpenAI · released 2026-03-17
textvision
Context
400K
Max out
128K
Input /1M
$0.75
Output /1M
$4.5
Cached /1M
$0.075
Scores
10 · 6 core
GPT-5.5
OpenAI · released 2026-04-23
textvision
Context
1M
Max out
128K
Input /1M
$5
Output /1M
$30
Cached /1M
$0.5
Scores
21 · 10 core
Quality

Benchmark matrix

BenchmarkGPT-5.4 miniGPT-5.5ΔEdge
Reported by both · 9
GPQA Diamond88% 90.7% epoch-2.7 ptGPT-5.5
FrontierMath Tier 4 v2 (Epoch AI run)9.8% epoch72.5% epoch-62.7 ptGPT-5.5
FrontierMath Tiers 1-3 v2 (Epoch AI run)51.2% epoch85.3% epoch-34.1 ptGPT-5.5
OSWorld-Verified72.1% 78.7% -6.6 ptGPT-5.5
OTIS Mock AIME 2024-2025 (Epoch AI run)88.9% epoch84.4% epoch+4.5 ptGPT-5.4 mini
SimpleQA Verified29.4% epoch63% epoch-33.6 ptGPT-5.5
SWE-Bench Pro54.4% 58.6% -4.2 ptGPT-5.5
Terminal-Bench 2.060% 82.7% -22.7 ptGPT-5.5
Toolathlon42.9% 55.6% -12.7 ptGPT-5.5
Only GPT-5.4 mini reports · 0
GPT-5.4 mini reports nothing GPT-5.5 does not.
Only GPT-5.5 reports · 12
AIME 2026not reported100% matharena
LMArena Elonot reported1465.6
ARC-AGI-1 (Verified)not reported95%
ARC-AGI-2 (Verified)not reported85%
BrowseCompnot reported84.4%
FrontierMath Tier 1-3not reported51.7%
FrontierMath Tier 4not reported35.4%
GDPval (wins or ties)not reported84.9%
MCP Atlasnot reported75.3%
MMMU-Pro (no tools)not reported81.2%
MMMU-Pro (with tools)not reported83.2%
Tau2-bench Telecomnot reported98%
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is GPT-5.4 mini minus GPT-5.5 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing

Side by side

SpecGPT-5.4 miniGPT-5.5Edge
Context window400K tokens1M tokensGPT-5.5
Max output128K tokens128K tokensTie
Input price / 1M$0.75 $5 GPT-5.4 mini
Output price / 1M$4.5 $30 GPT-5.4 mini
Cached input / 1M$0.075 $0.5 GPT-5.4 mini
Input + output / 1M
Lower is cheaper. List prices; batch, tool and regional fees excluded.
$5.25$35GPT-5.4 mini
Modalitiestext · visiontext · visionTie
Released2026-03-172026-04-23
Cited benchmark scores1021
Reliability

OpenAI status

All providers
More matchups

GPT-5.4 mini vs …

Built by Respan
Which one wins on your data?

Public benchmarks are a starting point. Run GPT-5.4 mini and GPT-5.5 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.