OpenAI

GPT-5 Nano — Benchmark Scores, Pricing & Performance Analysis

BUDGETOpenAI
Chatbot Arena ELO
1200
Output Speed
250 tok/s
Input Cost
$0.05/1M
Output Cost
$0.40/1M
Context Window
400K
Max Output
128K
Knowledge Cutoff
May 2024
Accepts
text, image, file

GPT-5 Nano by OpenAI demonstrates fast output speed, competitive pricing. View detailed benchmark data including scores across coding, math, reasoning, speed, and cost metrics.

Specifications last checked against the provider on . Benchmark scores carry their own source, linked beside each number.

General Benchmarks

MMLU-Pro
62.0%58th of 66

Massive Multitask Language Understanding professional benchmark

HuggingFace Leaderboard
Chatbot Arena ELO
120067th of 74

Human preference ELO from blind head-to-head votes

LMSYS / HuggingFace

Coding Benchmarks

Website Arena ELO
111380th of 108

Human preference ELO for building a working web page from a brief

Design Arena (via OpenRouter)
LiveCodeBench
30.0%56th of 66

Live competitive programming benchmark

livecodebench.github.io
Design Arena ELO
107194th of 120

Human preference ELO for building things — websites, UI, games, charts, SVG

Design Arena (via OpenRouter)
HumanEval+
72.0%58th of 66

Code generation correctness with extended tests

Papers
SWE-bench Verified
25.0%56th of 67

Real-world software engineering task resolution

swebench.com

Math Benchmarks

GSM8K
85.0%57th of 66

Grade school math word problems

Papers
MATH
65.0%57th of 66

Competition mathematics problem solving

Papers

Reasoning Benchmarks

ARC-AGI
15.0%48th of 66

Abstraction and Reasoning Corpus for general intelligence

arcprize.org
GPQA Diamond
70.9%87th of 140

Graduate-level science Q&A by domain experts

Papers
τ²-Bench Airline
47.6%74th of 93

Multi-turn service agent making tool calls under strict policy constraints

OpenRouter (measured)

Speed Benchmarks

Time to First Token
80 ms2nd of 92

Latency before first token arrives

Aggregated
Output Speed
250 tok/s4th of 94

Tokens generated per second

Aggregated

Cost Benchmarks

Output Cost
$0.40143rd of 413

Cost per 1M output tokens

OpenRouter API
Cached Input Cost
$0.015th of 136

Cost per 1M cached input tokens — the price that actually applies to a long agent conversation

OpenRouter API
Input Cost
$0.0588th of 413

Cost per 1M input tokens

OpenRouter API

Context Benchmarks

Context Length
400K90th of 451

Maximum context window size

OpenRouter API

Available from 28 providers

The same model costs different amounts depending on who serves it. You can bring your own key for 4 of these — connect it here.

ProviderInput / 1MOutput / 1MCached inContextYour key
Jiekou.AI$0.045$0.36—400K—
Poe$0.045$0.36$0.0045400K—
Abacus$0.05$0.4$0.005400K—
Azure$0.05$0.4$0.01400KSupported
Azure Cognitive Services$0.05$0.4$0.01400K—
Cloudflare AI Gateway$0.05$0.4$0.005128K—
Databricks$0.05$0.4$0.005400K—
DevPass (LLM Gateway)$0.05$0.4$0.005400K—
Eden AI$0.05$0.4$0.005400K—
FastRouter$0.05$0.4$0.005400K—
Impossibl$0.05$0.4$0.005400K—
Kilo Gateway$0.05$0.4$0.005400K—
LLM Gateway$0.05$0.4$0.005400K—
Merge Gateway$0.05$0.4$0.005400K—
NanoGPT$0.05$0.4$0.005400K—
NEAR AI Cloud$0.05$0.4$0.005400K—
Neon$0.05$0.4$0.005400K—
Ofox$0.05$0.4$0.01400K—
OpenAI$0.05$0.4$0.005400KSupported
OpenCode Zen$0.05$0.4$0.005400K—
OpenRouter$0.05$0.4$0.005400KSupported
OrcaRouter$0.05$0.4$0.005400K—
Pioneer$0.05$0.4$0.005400K—
SAP AI Core$0.05$0.4$0.005400K—
Vercel AI Gateway$0.05$0.4$0.005400KSupported
Requesty$0.055$0.44$0.0055200K—
Cortecs$0.06$0.439$0.019400K—
Snowflake Cortex———400K—

GPT-5 Nano — Benchmark Scores Overview

Scores normalized to percentage scale for visual comparison. ELO scores mapped to 0-100 range (1100-1500).

GPT-5 Nano — Frequently Asked Questions

How intelligent is GPT-5 Nano?

GPT-5 Nano scores 1200 on the Chatbot Arena ELO rating, making it an entry-level AI model. This score is based on blind head-to-head human preference voting.

How much does GPT-5 Nano cost?

GPT-5 Nano costs $0.05 per 1M input tokens and $0.40 per 1M output tokens. This makes it one of the more affordable models.

How fast is GPT-5 Nano?

GPT-5 Nano generates output at 250 tokens per second, which is very fast compared to other models. The time to first token is 80 ms.

How good is GPT-5 Nano at coding?

GPT-5 Nano achieves 25.0% on SWE-bench Verified, demonstrating basic real-world software engineering capability. This benchmark tests the model's ability to resolve actual GitHub issues.

How good is GPT-5 Nano at math and reasoning?

GPT-5 Nano scores 65.0% on the MATH benchmark (competition-level mathematics). It also achieves 70.9% on GPQA Diamond, a graduate-level science reasoning benchmark.

What is the context window of GPT-5 Nano?

GPT-5 Nano has a context window of 400K tokens. This determines how much text, conversation history, and code the model can process in a single request.

Who created GPT-5 Nano?

GPT-5 Nano was created by OpenAI. It is classified as a budget model in our catalogue.

Is GPT-5 Nano open source?

No, GPT-5 Nano is a proprietary model. It is available through OpenAI's API and compatible providers.