Leaderboard / OpenAI
GPT-4.1
OpenAI 2025-04-14 CLOSED Proprietary
Verdict
GPT-4.1 is behind the leaders: #250 of 406 by SIscope score. Strongest at following instructions, weaker at math. Mid-range on price.
SIscope score
36
#250 of 406
Human-vote rank
#175
of 419 · score 1383
Price per 1M tokens, blended
$3.5
$2 in · $8 out
Context window
1M
1,047,576 tokens
Shape
Each spoke is a category, 0 at the centre, 100 at the edge.
Category scores
Percentiles: 90 means better than about 90% of tested models.
What goes into each category · Compare with DeepSeek V4.1 Flash
Human-vote score over time
Text arena, recent updates.
Independent test runs
Run by Epoch AI on the same harness for every model. Bars show the share of questions answered correctly.
GPQA Diamond67%
PhD-level science multiple choice questions
Better than 39% of 333 tested models. Run Apr 14, 2025.
SimpleQA Verified31%
Short factual questions; rewards not making things up
Better than 27% of 100 tested models. Run Aug 31, 2026.
OTIS Mock AIME 2024-2538%
Competition maths problems in the style of AIME
Better than 28% of 313 tested models. Run Apr 14, 2025.
FrontierMath Tiers 1-36%
Very hard, unpublished research-style maths problems
Better than 6% of 122 tested models. Run Aug 27, 2026.
SWE-bench Verified49%
Fixing real GitHub issues in Python repositories
Better than 3% of 35 tested models. Run Feb 8, 2026.
Human-vote arenas
People compare two anonymous answers and pick the better one. Scores work like chess ratings.
| Arena | Rank | Score | Range (95%) | Votes | Updated |
|---|---|---|---|---|---|
| Text arena Everyday chat, blind human votes | #175 of 419 | 1383 | 1379–1387 | 50,298 | Oct 28, 2026 |
| WebDev arena Building web apps, blind human votes | Not on this board yet | ||||
| Vision arena Questions about images, blind human votes | #88 of 167 | 1210 | 1204–1217 | 40,998 | Oct 19, 2026 |
| Agent arena Real multi-step agent sessions | Not on this board yet | ||||
Source: Arena (formerly LMArena) leaderboard dataset, CC BY 4.0. Fetched Nov 4, 2026.
Price
Input
$2
per 1M tokens
Output
$8
per 1M tokens
Blended
$3.5
3 in : 1 out, per 1M
Context window
1M
max tokens per request
Speed
Coming soon
our own tests
List price from the OpenRouter models API (openai/gpt-4.1), Nov 4, 2026. Providers can charge differently.