GPT-5 nano is a streamlined variant of OpenAI's GPT-5 model, designed to deliver advanced multimodal reasoning and contextual understanding with significantly reduced computational overhead. It serves as an efficient alternative for developers and enterprises prioritizing fast inference and cost-effectiveness while retaining key features of the full GPT-5 system.
GPT-5 nano supports a large input context size of up to 400K tokens, matching GPT-5 full scale, enabling it to handle extensive documents and multimodal inputs such as text-to-text and image-to-text tasks efficiently.
GPT-5 nano inherits the advanced transformer framework of GPT-5 with optimized attention and efficient utilization of sparsity and mixture-of-experts layers tuned for lightweight operation. It balances architectural scale to sustain high throughput and lower compute costs while focusing on core reasoning and multimodal processing capabilities.
VS GPT-5 mini: GPT-5 nano focuses more on the fastest execution and lowest cost with basic multimodal support while GPT-5 mini balances speed and reasoning depth, supporting some expanded workflows with slightly higher pricing.
VS GPT-4o: GPT-5 nano significantly outperforms GPT-4o in reasoning accuracy, multimodal capabilities, and hallucination reduction, while maintaining much lower latency and cost compared to GPT-4o’s heavier but simpler model design.
VS OpenAI o3: GPT-5 nano provides more reliable fact-based answers and advanced reasoning than o3, with specialized alignment and safety mechanisms, delivering highly cost-efficient multimodal AI suitable for real-time applications.
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
import requests r = requests.post( "https://api.aimlapi.com/v1/chat/completions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "openai/gpt-5-nano-2025-08-07", "messages": [ { "role": "user", "content": "Hello!" } ] }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/chat/completions", { method: "POST", headers: { Authorization: `Bearer ${process.env.AIMLAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "openai/gpt-5-nano-2025-08-07", "messages": [ { "role": "user", "content": "Hello!" } ] }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/chat/completions \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"openai/gpt-5-nano-2025-08-07","messages":[{"role":"user","content":"Hello!"}]}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
| Cached input |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Intelligence | 13 | Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Math | 83.7 | Composite score across standardised mathematics evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
GPT-5 Nano This page | Reasoning + agents | |||
| Reasoning + agents | ||||
| Balanced coding + agents | ||||
| Long-context, multimodal & agentic workflows | ||||
| Reasoning + agents |
GPT-5 Nano has a 400,000 tokens context window and can return up to 128,000 tokens.
GPT-5 Nano takes image, text as input and returns text.
Use openai/gpt-5-nano-2025-08-07 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
GPT-5 Nano became available on September 30, 2025.
GPT-5 Nano is priced at input $0.065 / 1M tokens, output $0.52 / 1M tokens, cached input $0.0065 / 1M tokens.
Yes, GPT-5 Nano can stream responses as they are generated.
Yes, GPT-5 Nano accepts image input alongside text.
GPT-5 Nano was built by OpenAI.
It is the most lightweight variant of GPT-5, built for ultra-low latency use cases.
Yes, it supports function calling and parallel tool calls.
Yes, GPT-5 Nano supports structured output generation.
