跳转到主要内容
GPT-5.3-chat-latest is OpenAI’s latest chat model released on March 3, 2026, the GPT-5.3 Instant model used in ChatGPT. Compared to its predecessor, it features significant improvements in accuracy, context understanding, and conversational fluency. API Yi connects through official direct channels for reliable and stable service.
API Yi now supports GPT-5.3-chat-latest via official direct connection, compatible with OpenAI format, plug and play.

​
Key Advantages

Massive Context

400K token context window, 3x the previous generation, easily handles ultra-long documents

Lower Hallucination

26.8% reduction in hallucination rate (with web search), 19.7% (knowledge-only)

Natural Conversation

Fewer unnecessary refusals and preachy responses, smoother and more natural dialogue

Official Direct Connection

OpenAI official API transparent forwarding, same quality and stability as official

​
Model Information

ParameterValue
Model Namegpt-5.3-chat-latest
Context Window128,000 tokens
Max Output16,384 tokens
Knowledge CutoffAugust 31, 2025
Input FormatText + Image
Output FormatText
Streaming✅ Supported
Function Calling✅ Supported
Structured Output✅ Supported

​
Pricing

ItemPrice (per million tokens)
Input$1.75
Cached Input$0.175
Output$14.00
Pricing matches OpenAI official rates. Top-up bonuses available for better value. See Pricing for details.

​
How to Use

​
Endpoint

https://api.apiyi.com/v1/chat/completions

​
Basic Call

curl -X POST "https://api.apiyi.com/v1/chat/completions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.3-chat-latest",
    "messages": [
      {"role": "user", "content": "Hello, tell me about your capabilities"}
    ]
  }'

​
Streaming Output

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.apiyi.com/v1"
)

response = client.chat.completions.create(
    model="gpt-5.3-chat-latest",
    messages=[
        {"role": "user", "content": "Write a short essay about the future of AI"}
    ],
    stream=True
)

for chunk in response:
    if chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="")

​
Multimodal Input (Vision)

GPT-5.3-chat-latest supports image input for understanding and analyzing visual content: 
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.apiyi.com/v1"
)

response = client.chat.completions.create(
    model="gpt-5.3-chat-latest",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Describe the content of this image"},
                {
                    "type": "image_url",
                    "image_url": {
                        "url": "https://example.com/image.png"
                    }
                }
            ]
        }
    ]
)

print(response.choices[0].message.content)

​
Function Calling

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://api.apiyi.com/v1"
)

tools = [
    {
        "type": "function",
        "function": {
            "name": "get_weather",
            "description": "Get weather information for a specified city",
            "parameters": {
                "type": "object",
                "properties": {
                    "city": {
                        "type": "string",
                        "description": "City name"
                    }
                },
                "required": ["city"]
            }
        }
    }
]

response = client.chat.completions.create(
    model="gpt-5.3-chat-latest",
    messages=[
        {"role": "user", "content": "What's the weather like in Beijing today?"}
    ],
    tools=tools,
    tool_choice="auto"
)

print(response.choices[0].message)

​
Comparison with Other GPT Models

FeatureGPT-5.3 ChatGPT-5.2GPT-5
Context Window128K128K400K
Max Output16K16K128K
Input Price$1.75/M$1.75/M$1.25/M
Output Price$14.00/M$14.00/M$10.00/M
Hallucination Improvement✅ 26.8% reduction--
Conversation Naturalness✅ Significantly improvedGoodGood
Image Input✅✅✅
Function Calling✅✅✅

​
FAQ

GPT-5.3 Chat is the latest Instant model used in ChatGPT. Compared to GPT-5.2, it reduces hallucination rates by 26.8% (with web search), features more natural conversation style, and eliminates unnecessary refusals and preachy responses.
GPT-5.3 Chat is ideal for everyday conversations, customer service, content creation, and Q&A scenarios requiring natural and fluent dialogue. For coding tasks, consider using the GPT-5.3-Codex model.
Yes, GPT-5.3-chat-latest supports text and image input for understanding and analyzing visual content. However, audio and video inputs are not supported.
The maximum output is 16,384 tokens. If output is truncated, check the finish_reason field and adjust the max_tokens parameter accordingly.
When your requests contain large repeated prefix content (such as system prompts), OpenAI automatically caches these. Cached input is billed at $0.175/M tokens, just 1/10 of the regular input price.
Yes, API Yi uses OpenAI’s official API transparent forwarding, ensuring the same service quality and stability as the official API.

Text Generation Guide

Complete guide to Chat Completions API

Model Information

View all available models and pricing

OpenAI SDK Guide

OpenAI SDK configuration and usage tutorial

Use Cases

Explore various use case configurations