Speed
Output tokens / second
GPT-5 Nano overview
GPT-5 Nano by OpenAI
It supports extensive context processing and key NLP tasks such as summarization and classification, making it ideal for developers and enterprises needing fast, affordable, and versatile AI across text-to-text and image-to-text workflows.
GPT-5 Nano specs
- Model ID
openai/gpt-5-nano-2025-08-07- Developer
- OpenAI
- Context window
- 400,000 tokens
- Max output
- 128,000 tokens
- Input
- Text
- Output
- Text, Image
- API
- OpenAI-compatible, base URL
https://api.aimlapi.com/v1
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.
Context Window and Token Capacity
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.
Performance Benchmarks
- Speed & Latency: Optimized for low-latency inference with trade-offs favoring faster response times over the deepest reasoning layers of full GPT-5.
- Accuracy: Retains strong few-shot learning, multimodal understanding, and factual correctness, though with slightly less complexity handling than GPT-5 and GPT-5 mini.
- Multilingual support: Comprehensive, leveraging GPT-5’s expanded language capabilities.
Architecture Highlights
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.
Core Features & Capabilities
- Model Scale: Smaller parameter count than GPT-5 and mini, designed for speed and resource efficiency without substantial sacrifices in contextual understanding or multimodal tasks.
- Multimodality: Supports text-to-text and vision (image-to-text) input modalities through the API. Audio, video, and code input functionalities remain targeted for future expansions in the unified GPT-5 framework.
- Reasoning: Capable of stepwise logical reasoning and complex problem solving, though optimized for faster execution over the most compute-intensive scenarios.
- Fine-Tuning & Adaptability: Enables flexible customization for domain-specific tasks and enterprise needs.
- Bias & Safety: Implements advanced alignment, bias mitigation, and safety features consistent with GPT-5’s standards.
Use Cases & Applications
- Fast multimodal content understanding and generation in cost-sensitive environments.
- Scalable deployment for lightweight software engineering support, including code suggestions and debugging.
- Real-time large-scale document analysis with image context integration.
- Educational tools and research assistants requiring concise and accurate multi-step instruction processing.
Comparison with Other Models
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.
Use cases
- Fast multimodal content understanding and generation in cost-sensitive environments.
- Scalable deployment for lightweight software engineering support, including code suggestions and debugging.
- Real-time large-scale document analysis with image context integration.
- Educational tools and research assistants requiring concise and accurate multi-step instruction processing.
Capabilities
- Function calling
- Streaming
- Structured outputs
- Reasoning
GPT-5 Nano benchmarks by 
Artificial Analysis
See how GPT-5 Nano compares in generation speed. Data retrieved Nov 3, 2026.
Benchmark results
Compare 17 models available on AI/ML API across 5 independent benchmarks.
Humanity's Last Exam
Expert-written questions across academic subjects.
GPQA Diamond
Graduate-level science questions written to resist web search.
AA-LCR
Reasoning across long documents of about 100k tokens.
IFBench
Following precise instructions about the shape of the answer.
τ²-Bench Telecom
Customer support agent tasks with tool calls.
GPT-5 Nano API pricing
| Usage | Price | Unit |
|---|---|---|
| Input | $0.05 | 1M tokens |
| Output | $0.4 | 1M tokens |
| Cached input | $0.005 | 1M tokens |
Prices from the AI/ML API price list, checked on Nov 4, 2026.
GPT-5 Nano vs other models: token prices
AI/ML API input and output rates for 5 models, per million tokens.
| Model | Input | Output |
|---|---|---|
 GPT-5 Nano | $0.05 | $0.4 |
| GPT-6 Luna | $0.1 | $0.5 |
| GPT-6.1 Terra | $1 | $5 |
| GPT-6.1 Sol | $2 | $10 |
| GPT-6.1 Terra Pro | $2.12 | $10.58 |
How to use the GPT-5 Nano API
View docs
Build with your agent
Give your agent the model details, code and documentation.
Paste it into your coding agent to build the integration.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl '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": "Write a haiku about the ocean."
}
]
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "openai/gpt-5-nano-2025-08-07",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}
response = requests.post("https://api.aimlapi.com/v1/chat/completions", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = 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": "Write a haiku about the ocean."
}
]
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Models similar to GPT-5 Nano
Related articles
GPT-5 Nano FAQ
What is the context window of GPT-5 Nano?
GPT-5 Nano has a 400,000 tokens context window and can return up to 128,000 tokens.
What inputs and outputs does GPT-5 Nano support?
GPT-5 Nano takes image, text as input and returns text.
What is the model ID for GPT-5 Nano?
Use openai/gpt-5-nano-2025-08-07 as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
When was GPT-5 Nano released?
GPT-5 Nano became available on September 30, 2025.
How much does GPT-5 Nano cost?
GPT-5 Nano costs $0.05 input and $0.4 output per 1M tokens on AI/ML API, billed as you go with no subscription.
Does GPT-5 Nano support streaming?
Yes, GPT-5 Nano can stream responses as they are generated.
Can GPT-5 Nano read images?
Yes, GPT-5 Nano accepts image input alongside text.
Who created GPT-5 Nano?
GPT-5 Nano was built by OpenAI.
What is GPT-5 Nano designed for?
It is the most lightweight variant of GPT-5, built for ultra-low latency use cases.
Does GPT-5 Nano support function or tool calling?
Yes, it supports function calling and parallel tool calls.