openai/gpt-5-nano-2025-08-07

openai/gpt-5-nano-2025-08-07
VisionToolsJSONReasoning
by openai

OpenAI GPT-5 Nano: 400K context, 83.7 AA Math, text/image/file input, $0.05/M input tokens via OrcaRouter.

ctx400K tokens
Max output128K
Inputtext + image + file
Outputtext
p50 TTFT5.00 s
INPUT$0.05/ 1M tokens
OUTPUT$0.40/ 1M tokens
p50 TTFT5.00 s7d
p95 TTFT10.00 s7d
TRAFFIC754.6Ktokens / 7d

OpenAI GPT-5 Nano (2025-08-07 release) is a smaller, cost-optimized member of the GPT-5 family. It offers a context window of 400,000 tokens—enough to process entire long documents or book-length…

What is OpenAI GPT-5 Nano?

Who is GPT-5 Nano designed for?

Why access GPT-5 Nano through OrcaRouter?

When should you consider an alternative to GPT-5 Nano?

Code samples

Call from any SDK

OpenAI-compatible — keep the SDK you already use

  • OpenAI SDKhttps://api.orcarouter.ai/v1
from openai import OpenAI

client = OpenAI(
    base_url="https://api.orcarouter.ai/v1",
    api_key="$ORCAROUTER_API_KEY",
)

response = client.chat.completions.create(
    model="openai/gpt-5-nano-2025-08-07",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Supported parameters

  • include_reasoning
  • max_completion_tokens
  • max_tokens
  • reasoning
  • response_format
  • seed
  • stream
  • structured_outputs
  • tool_choice
  • tools

Pricing

Input / 1M tokens$0.050
Output / 1M tokens$0.400
Cache read / 1M$0.0050
CurrencyUSD

Cost calculator

Tokens / month10MM
Input share70%%
Estimated / month $1.55 · With prompt caching $1.39

Estimate based on list price

Token & cost estimator

Input tokens: 20Cost per request: $0.000201

Estimate only — actual token counts depend on the provider's tokenizer.

Performance

p50 TTFT
5.00 s
Output speed
554 tok/s
p95 TTFT
10.00 s
Error rate
0%

Public benchmarks

20.3
AA Coding
Better than 19% of models compared
#85 of 106
26.8
AA Intelligence
Better than 27% of models compared
#79 of 110
83.7
AA Math
Better than 73% of models compared
#21 of 81
AIME 2025
83.7
GPQA Diamond
67.6
Humanity's Last Exam
8.2
IFBench
67.6
LiveCodeBench
78.9
Long-Context Recall
41.7
MMLU-Pro
78.0
SciCode
36.6
TerminalBench Hard
12.1
τ²-Bench
36.5
Source: artificialanalysis.ai

How it compares

openai/gpt-5-nano-2025-08-07GPT-5.2 ProGPT-5.4 ProGPT-5.5
Input $/M$0.05$21.00$60.00$5.00
Output $/M$0.40$168.00$270.00$30.00
Context400K400K1.1M
Quality4/1010/1010/1010/10
Compare side-by-sideCompare side-by-sideCompare side-by-sideCompare side-by-side

FAQ

What is the cost per million tokens for GPT-5 Nano?
Input tokens: $0.05 per 1M tokens. Output tokens: $0.40 per 1M tokens. These are the provider rates passed through by OrcaRouter with zero markup.
What is the context window of GPT-5 Nano?
400,000 tokens total for input (including text, images, and file tokens). Maximum output is 128,000 tokens per generation.
What are GPT-5 Nano’s main strengths?
Very long context (400K), multi-modal input (text/image/file), large output (128K), strong math reasoning (83.7 AA Math), and low input cost.
How does GPT-5 Nano compare to larger GPT-5 models?
GPT-5 Nano is smaller and cheaper but doesn't achieve the highest scores on non-math benchmarks. Larger Pro models have higher accuracy at higher cost.
Does OrcaRouter guarantee zero markup on GPT-5 Nano pricing?
Yes, OrcaRouter bills at the OpenAI provider rate with zero markup. You pay exactly $0.05/1M input and $0.40/1M output.
What data handling policy applies to GPT-5 Nano via OrcaRouter?
OpenAI’s API data policy applies—your data is not used for training. OrcaRouter does not retain prompts beyond standard billing logs. See OrcaRouter’s privacy policy for details.
How do I call GPT-5 Nano using an OpenAI-compatible API?
Set base URL to https://api.orcarouter.ai/v1, model ID "openai/gpt-5-nano-2025-08-07", and use standard chat completions parameters. Replace your API key with an OrcaRouter key.
Can GPT-5 Nano process images and files?
Yes, it supports text, image, and file inputs. Images and files are tokenized and count toward the 400K context limit.
What is the maximum output length?
Up to 128,000 tokens. Use the max_tokens parameter to control output length.
Is GPT-5 Nano good for math problems?
Yes, its AA Math score of 83.7 indicates strong mathematical reasoning, suitable for arithmetic, algebra, and word problems.

Embed this badge

openai/gpt-5-nano-2025-08-07$0.05/M in5000ms p50via OrcaRouter
HTML <a href="https://www.orcarouter.ai/models/openai/gpt-5-nano-2025-08-07" target="_blank"> <img src="https://www.orcarouter.ai/embed/openai/gpt-5-nano-2025-08-07.svg" alt="openai/gpt-5-nano-2025-08-07 on OrcaRouter" /> </a>
Markdown [![openai/gpt-5-nano-2025-08-07](https://www.orcarouter.ai/embed/openai/gpt-5-nano-2025-08-07.svg)](https://www.orcarouter.ai/models/openai/gpt-5-nano-2025-08-07)

Model card as data

GET /api/public/models/openai/gpt-5-nano-2025-08-07Open
Machine-readable:/llms.txt/llms-full.txt