GPT-4.1 (2025-04-14)
Apr 2025
GPT-4.1 (2025-04-14)Compare inon
Benchmarks of GPT-4.1 (2025-04-14)36
BenchmarkByScoreRank
Knowledge14
Reasoning1
Math4
Coding12Leads on 1
Legacy · saturated or no longer reported5
No benchmark matches
Ranking?
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperMuse Sparkunstated×7.1-43%$1.25 / $4.25Only n=1 · Meta / Chat Completions+1
BetterClaude Opus 4.8max×8.6+186%$4.08
OpenKimi K2.6unstated×7.1-51%$0.95 / $4Context: 1M → 256K
CheaperGemini 3 Flash Previewunstated×6.5-68%$0.50 / $3
CheaperGLM 5.1unstatedOpen×6.1-39%$1.40 / $4.40Context: 1M → 203K+1ModelScoreCostPer taskWhat changes
Muse Sparkunstated×7.1-43%$1.25 / $4.25Only n=1 · Meta / Chat Completions+1Kimi K2.6unstatedOpen×7.1-51%$0.95 / $4Context: 1M → 256KGemini 3 Flash Previewunstated×6.5-68%$0.50 / $3GLM 5.1unstatedOpen×6.1-39%$1.40 / $4.40Context: 1M → 203K+1GPT-5.4 Mini (2026-03-17)high×5.1-52%$0.75 / $4.50Context: 1M → 400KModelScoreCostPer taskWhat changes
Claude Opus 4.8max×8.6+186%$4.08GPT-5.4 Pro (2026-03-05)×9.1+1,829%$30 / $180No strict JSONGPT-5.4 (2026-03-05)xhigh×8.6+61%$2.50 / $15Claude Opus 4.7xhigh×7.9+186%$5 / $25Claude Opus 4.6max×7.4+186%$5 / $25ModelScoreCostPer taskWhat changes
Kimi K2.6unstatedOpen×7.1-51%$0.95 / $4Context: 1M → 256KGLM 5.1unstatedOpen×6.1-39%$1.40 / $4.40Context: 1M → 203K+1Kimi K2 ThinkingunstatedOpen×3.9-76%$0.47 / $2Context: 1M → 256K+1GLM 5unstatedOpen×3-56%$1 / $3.20Context: 1M → 203K+1
Qwen3 235B A22B Thinking 2507unstatedOpen+54%-79%$0.23 / $2.30Context: 1M → 262K+2ModelScoreCostPer taskWhat changes
GPT-5.4 Pro (2026-03-05)×9.1—$30 / $180No strict JSONGPT-5.4 (2026-03-05)xhigh×8.6—$2.50 / $15GPT-5.2 (2025-12-11)medium×6.7—$1.75 / $14Context: 1M → 400KGPT-5.4 Mini (2026-03-17)high×5.1—$0.75 / $4.50Context: 1M → 400KGPT-5 (2025-08-07)medium×4.9—$1.25 / $10Context: 1M → 400KGPT-5.1 (2025-11-13)medium×4.9—$1.25 / $10Context: 1M → 400KGPT-5.4 Nano (2026-03-17)high×4.7—$0.20 / $1.25Context: 1M → 400KGPT-5 Mini (2025-08-07)medium×3.7—$0.25 / $2Context: 1M → 400KAll OpenAI models →History
17 versionsBenchmarks
all categoriesChat & Knowledge
Arena Text · Chinese1,381.9#114
Arena Text · English1,392.3#100
Arena Text · French1,381.9#98
Arena Text · German1,381.1#79
Arena Text · instruction following1,366.8#98
Arena Text · Japanese1,319#80
Arena Text · Korean1,338.6#81
Arena Text · non-English1,369.7#99
Arena Text · Polish1,388.8#75
Arena Text · Russian1,377#96
Arena Text · Spanish1,375.8#99
SimpleQA Verified v131.1#56
MATH Level 583#9
FrontierMath5.5#30
Chess Puzzles3#66
Arena Text · overall · full cohort1,382.9#101
GPQA Diamond · Epoch66.9#85
OTIS Mock AIME 2024 202538.3#97
Coding
SWE-rebench · all · 2025-01-01 to 2025-08-0130.1#1
SWE-rebench · Go · 2025-01-01 to 2025-08-010#1
SWE-rebench · Java · 2025-01-01 to 2025-08-010#1
SWE-rebench · Rust · 2025-01-01 to 2025-08-010#1
SWE-rebench · Python · 2025-01-01 to 2025-08-0130.1#1
SWE-rebench · TypeScript · 2025-01-01 to 2025-08-010#1
SWE-rebench · all · 2025-05-01 to 2025-12-0129.7#2
SWE-rebench · Go · 2025-05-01 to 2025-12-010#1
SWE-rebench · Java · 2025-05-01 to 2025-12-010#1
SWE-rebench · Rust · 2025-05-01 to 2025-12-010#1
SWE-rebench · Python · 2025-05-01 to 2025-12-0129.7#2
SWE-rebench · TypeScript · 2025-05-01 to 2025-12-010#1
SWE-bench Verified · Epoch48.5#23
Agents & Tool Use
BFCL v454#9
BFCL v4 · prompt mode39.4#5
Writing & Language
FAQ
How much does GPT-4.1 (2025-04-14) cost?›
$2 per 1M input tokens and $8 per 1M output tokens.
Is GPT-4.1 (2025-04-14) free to use?›
There is no free API; the cheapest paid price is $2 per 1M input tokens.
How good is GPT-4.1 (2025-04-14), and what is it best at?›
#100 of 188 in Chat & Knowledge, #23 of 24 in Coding, #9 of 39 in Agents & Tool Use, #87 of 187 in Writing & Language.
Is GPT-4.1 (2025-04-14) better than Qwen 3.5 Flash?›
On FrontierMath, GPT-4.1 (2025-04-14) scores 5.5 vs 6.2 — about the same.
How can I use GPT-4.1 (2025-04-14)?›
Through the OpenAI API.
What is the context window of GPT-4.1 (2025-04-14)?›
1M tokens.
When was GPT-4.1 (2025-04-14) released? Is there a newer version?›
Released 2025-04-14. It has been superseded by GPT-6 Luna.
Can GPT-4.1 (2025-04-14) read images?›
It accepts text and images as input.
What is GPT-4.1 (2025-04-14)?›
GPT-4.1 (2025-04-14) is a language model from OpenAI, released Apr 2025.
What are the best alternatives to GPT-4.1 (2025-04-14)?›
Muse Spark (cheaper), Claude Opus 4.8 (better), Kimi K2.6 (open weights).