Compare

Four models, one page

Pick the models you are actually choosing between. Every row is scaled within itself, so a bar means something next to its neighbours and nothing across rows.

Slot 1

Slot 2

Slot 3

Slot 4

Specification	GPT-4.1 MiniOpenAI	Claude Opus 5Anthropic
Released	Apr 14, 2025	Jul 24, 2026
Context window	1.05M	1M
Max output	33K	128K
Input / 1M	$0.4	$5
Output / 1M	$1.6	$25
Cost per task	$0.06	$2.25
Arena Elo	1,112	1,393
Serving providers	2	5
Parameters	Undisclosed	Undisclosed
Licence	Proprietary	Proprietary
Capabilities	Tool use Vision	Reasoning Tool use Vision

Metrics side by side

Each row is scaled to the largest value in that row — bars compare within a row, never across rows

GPT-4.1 Mini
Claude Opus 5

Intelligence Index
GPT-4.1 Mini
14.8
Claude Opus 5
63.1
Coding Index
GPT-4.1 Mini
20.2
Claude Opus 5
78.0
Agentic Index
GPT-4.1 Mini
1.8
Claude Opus 5
59.2
Output speed
GPT-4.1 Mini
40 t/s
Claude Opus 5
87 t/s
Context window
GPT-4.1 Mini
1.05M
Claude Opus 5
1M
Latency · lower is better
GPT-4.1 Mini
824ms
Claude Opus 5
1.12s
Blended price / 1M · lower is better
GPT-4.1 Mini
$0.7
Claude Opus 5
$10
Cost per task · lower is better
GPT-4.1 Mini
$0.06
Claude Opus 5
$2.25

Rows marked “lower is better” still draw a longer bar for a larger number — read the value, not just the length. Arena Elo is in the specification table above instead: it has no meaningful zero, so a bar would flatten the gaps.

View as table

Metric	GPT-4.1 Mini	Claude Opus 5
Intelligence Index	14.8	63.1
Coding Index	20.2	78.0
Agentic Index	1.8	59.2
Output speed	40 t/s	87 t/s
Context window	1.05M	1M
Latency · lower is better	824ms	1.12s
Blended price / 1M · lower is better	$0.7	$10
Cost per task · lower is better	$0.06	$2.25

Evaluation scores

Percentage correct on a common 0–100% scale

GPT-4.1 Mini
Claude Opus 5

GPQA Diamond
GPT-4.1 Mini
66.4%
Claude Opus 5
93.2%
Humanity's Last Exam
GPT-4.1 Mini
5.0%
Claude Opus 5
54.9%
SciCode
GPT-4.1 Mini
40.4%
Claude Opus 5
55.7%
τ²-bench
GPT-4.1 Mini
52.9%
Claude Opus 5
42.1%
Terminal-Bench Hard
GPT-4.1 Mini
7.6%
Claude Opus 5not measured
LiveCodeBench
GPT-4.1 Mini
48.3%
Claude Opus 5not measured
AA-LCR long context
GPT-4.1 Mini
45.3%
Claude Opus 5
75.7%
AIME 2025
GPT-4.1 Mini
46.3%
Claude Opus 5not measured

A missing bar means that evaluation was not run for that model — it is not a zero.

View as table

Evaluation	GPT-4.1 Mini	Claude Opus 5
GPQA Diamond	66.4%	93.2%
Humanity's Last Exam	5.0%	54.9%
SciCode	40.4%	55.7%
τ²-bench	52.9%	42.1%
Terminal-Bench Hard	7.6%	—
LiveCodeBench	48.3%	—
AA-LCR long context	45.3%	75.7%
AIME 2025	46.3%	—

Estimated from list pricing: 50K input tokens plus 80K output tokens for reasoning models (25K for non-reasoning).