BenchLeader

GPT-4.1 nano

Reasoning effort
Quality#302 of 333
36.9Index
Bottom quarter
−13.1 on the median
Price#31 of 249
$0.175per 1M tokens
Cheapest quarter
0.2× the median
Speed#30 of 215
148tokens/s
Top quarter
2.0× the median

22 cheaper ranked models score higher: Step 3.5 Flash (56.4, $0.150), GPT-5 nano (55.9, $0.138), MiMo-V2-Flash (54.0, $0.150) and 19 more.

As of 8 Nov 2026, GPT-4.1 nano by OpenAI scores 36.9 on the BenchLeader Index (593 ranked configurations), with its best results in human preference (48) and weakest in knowledge (20). API pricing: $0.100 per 1M input tokens and $0.400 per 1M output tokens ($0.175 blended at 3:1), list price on 8 Nov 2026. It streams about 150 tokens/s (Artificial Analysis). Released 14 Apr 2025.

Key facts (as of 8 Nov 2026)

  • GPT-4.1 nano API pricing: $0.100 per 1M input tokens and $0.400 per 1M output tokens ($0.175 blended at 3:1) (list price, 8 Nov 2026).
  • GPT-4.1 nano streams about 150 tokens/s (Artificial Analysis, 8 Nov 2026).
  • GPT-4.1 nano was released on 14 Apr 2025 (OpenAI).

In detail. GPT-4.1 nano is an OpenAI proprietary model, released 14 Apr 2025. Its best configuration ranks #559 of 593 on the BenchLeader Index at 36.9 ±4.4, in the lower half. The ± is the point: 109 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in human preference (48) and lowest in knowledge (20). At $0.175 per million tokens blended it is among the cheapest fifth of ranked models. Output speed of 148 tokens per second puts it in the fastest quarter, with a first answer in 1.1 s. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 6 Nov 2026.

Blended price
$0.175/M
$0.100 in · $0.400 out
via Artificial Analysis
Output speed
148 tok/s
measured by Artificial Analysis
First answer
1.12 s
first token 1.09 s
Context
1.0M
via models.dev
Full answer
6.94 s
median, reasoning included
Cost per task
–
Tokens per task
–
Released
14 Apr 2025
How it scores by categoryDashed line = average model (50). One step of 15 = one standard deviation.
  1. Overall index37
  2. Reasoning34
  3. Coding30
  4. Agents & tools38
  5. Maths46
  6. Knowledge20
  7. Instruction following47
  8. Human preference48
  9. Multimodal27
  10. Long context29

Get an email when OpenAI ships a new model

And once when it gets ranked, plus the weekly review on Mondays.

Choose providers and emails

Versions

OpenAI has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.

ModelReleasedIndexRank
GPT-5.4 Nanoxhigh17 Mar 202654.1#235
GPT-5 nano7 Aug 202555.9#192
GPT-4.1 nanothis page14 Apr 202536.9#559

Reasoning-effort configurations

The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.

EffortIndexRankSpeedFirst answerChat reply costCategories
high32.0#585148 tok/s1.12 s$0.0002Coding 28 · Knowledge 27 · Maths 21
not statedbest36.9#559148 tok/s1.12 s$0.0002Agents & tools 38 · Coding 30 · Human preference 48 · Instruction following 47 · Knowledge 20 · Long context 29 · Maths 46 · Multimodal 27 · Reasoning 34

How its index has moved

8 Sep 2026 to 8 Nov 2026
273544
  • not stated
  • high

The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.

Benchmark results

See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.

Reasoning

Benchmarkhighnot statedSourceTrend
GPQA Diamondeffort not stated–48.9%#231Epoch AI Benchmarking Hub
Arena Hard Promptseffort not stated–1333#234Arena (LMArena)
GPQA Diamond (AA)not in indexeffort not stated–51.2%#433Artificial Analysis
Humanity's Last Exam (AA)not in indexeffort not stated–3.8%#586Artificial Analysis
GPQA Diamond (Vals)not in index50.8%#117–Vals AI
Kagi LLM Benchmarkeffort not stated–33.3%#121Kagi LLM Benchmark
GPQA Diamond (HELM)not in indexeffort not stated–50.7%#37HELM Capabilities
ARC-AGI-1effort not stated–0.0%#259ARC Prize
ARC-AGI-2effort not stated–0.0%#245ARC Prize
CritPtnot in indexeffort not stated–0.0%#329Artificial Analysis
Chess Puzzlesnot in indexeffort not stated–0.0%#154Epoch AI Benchmarking Hub
LMCAnot in indexeffort not stated–5.5%#191Epoch AI Benchmarking Hub
DTBenchnot in indexeffort not stated–52.5%#207Epoch AI Benchmarking Hub

Coding

Benchmarkhighnot statedSourceTrend
SciCodeeffort not stated–25.9%#218SciCode
WeirdMLeffort not stated–19.0%#146WeirdML
Arena Codingeffort not stated–1373#226Arena (LMArena)
LiveCodeBench42.7%#124–Vals AI
Aider Polygloteffort not stated–8.9%#47Aider polyglot leaderboard
Terminal-Bench 2.1 (AA)not in indexeffort not stated–3.8%#217Artificial Analysis
Design Arena Websitenot in indexeffort not stated–970#171Design Arena
LiveCodeBench (AA)not in indexeffort not stated–32.6%#192Artificial Analysis

Maths

Benchmarkhighnot statedSourceTrend
OTIS Mock AIMEeffort not stated–28.9%#220Epoch AI Benchmarking Hub
MATH Level 5effort not stated–70.0%#40Epoch AI Benchmarking Hub
AIME (Vals)26.5%#75–Vals AI
MGSM69.3%#72–Vals AI
Omni-MATH (HELM)effort not stated–36.7%#35HELM Capabilities
MATH 500not in index80.2%#38–Vals AI
Arena Mathseffort not stated–1273#259Arena (LMArena)

Knowledge

Benchmarkhighnot statedSourceTrend
SimpleQA Verifiedeffort not stated–6.0%#97Epoch AI Benchmarking Hub
MMLU-Pro63.5%#129–Vals AI
LegalBench61.1%#140–Vals AI
CorpFin42.1%#119–Vals AI
TaxEval60.8%#124–Vals AI
MedQA68.2%#81–Vals AI
MMLU-Pro (HELM)not in indexeffort not stated–55.0%#53HELM Capabilities

Instruction following

Human preference

Benchmarknot statedSourceTrend
Arena Text1322#239Arena (LMArena)
Arena Creative Writingnot in index1307#214Arena (LMArena)
Arena Multi-turnnot in index1314#238Arena (LMArena)
Arena Longer Queriesnot in index1325#242Arena (LMArena)

Multimodal

Benchmarkhighnot statedSourceTrend
Arena Visioneffort not stated–1063#133Arena (LMArena)
MMMU-Proeffort not stated–40.1%#277Artificial Analysis
VISTAeffort not stated–26.6%#51Scale AI SEAL
MMMU-Pro (Vals)not in index55.1%#88–Vals AI
MortgageTaxnot in index52.8%#79–Vals AI

Long context

Benchmarknot statedSourceTrend
Fiction.LiveBench 120k18.8%#36Fiction.live
AA-LCR v1.120.3%#448Artificial Analysis

Composite

Safety & honesty

Benchmarknot statedSourceTrend
MASKnot in indexpossible contamination61.4%#35Scale AI SEAL

API pricing: where to run it

Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.

ProviderSpeedFirst tokenInput $/MOutput $/MContextQuantisation
Azure203 tok/s0.62 s$0.110$0.4401.0M–
Azure114 tok/s0.99 s$0.100$0.4001.0M–

What a task costs

Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.025 per 1M. Reasoning tokens are not modelled.

WorkloadTokens in / outCostWith cachingTime
Chat reply400 / 300$0.0002$0.00013.1 s
Summarize a 30-page report12,000 / 600$0.0014$0.00085.2 s
Code edit6,000 / 1,500$0.0012$0.000911.2 s
Agentic coding session60,000 / 4,000$0.0076$0.004228.1 s
Structured extraction2,000 / 200$0.0003$0.00022.5 s

More from OpenAI

New this month

See also

Embed or cite

Show where GPT-4.1 nano ranks on a model card or README. The badge redraws itself with every daily update, carries the data date and links back to this page. To see ranks like this on OpenRouter and Hugging Face model pages, get the Chrome extension.

[![GPT-4.1 nano on BenchLeader: #294 overall](https://www.benchleader.com/badge/gpt-4-1-nano.svg)](https://www.benchleader.com/models/gpt-4-1-nano)

Cite this

The ranking changes daily, so a reference carries the data date. APA style:

BenchLeader. (2026, November 8). GPT-4.1 nano on the BenchLeader Index (method v1.4) [Data set]. Retrieved November 8, 2026, from https://www.benchleader.com/models/gpt-4-1-nano

BibTeX:

@misc{benchleader_gpt_4_1_nano_20261108,
  author       = {{BenchLeader}},
  title        = {{GPT-4.1 nano on the BenchLeader Index (method v1.4)}},
  year         = {2026},
  month        = nov,
  howpublished = {\url{https://www.benchleader.com/models/gpt-4-1-nano}},
  note         = {Data as of 2026-11-08}
}

Every badge style, the embeddable top-10 widget and how to cite the Index

Questions people ask

What are GPT-4.1 nano's benchmark scores?
As of 8 Nov 2026, GPT-4.1 nano scores 36.9 ±4.4 on the BenchLeader Index, strongest in human preference (48) and weakest in knowledge (20). That rests on independent benchmark results only, and every score on this page links to its publisher.
How does GPT-4.1 nano rank?
GPT-4.1 nano is #559 of 593 ranked configurations on 8 Nov 2026. 109 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 69.4.
How much does GPT-4.1 nano cost per 1M tokens?
As of 8 Nov 2026, GPT-4.1 nano costs $0.100 per 1M input tokens and $0.400 per 1M output tokens, $0.175 blended (three input tokens per output token).
When was GPT-4.1 nano released?
OpenAI released GPT-4.1 nano on 14 Apr 2025. BenchLeader has results for it at 2 reasoning-effort settings.

Data as of 8 Nov 2026. Compare these configurations.

Cite as: BenchLeader, “GPT-4.1 nano: benchmarks, pricing, speed and rank”, data as of 8 Nov 2026, https://www.benchleader.com/models/gpt-4-1-nano.