# GPT-5.3 Codex - Core.Today AI API

> Codex model on the GPT-5.3 base, optimized for agentic coding harnesses and served via the OpenAI Responses API. 400K token context window, 128K max output tokens, cached inputs at a 90% discount.

- **Provider**: OpenAI
- **Model ID**: gpt-5.3-codex
- **Category**: LLM
- **Credits**: 15 per 1K tokens (avg)
- **Speed**: Medium
- **Quality**: Ultra

## Model Specifications
- **Context Window**: 400K tokens
- **Max Output**: 128K tokens
- **Training Cutoff**: Not published
- **Supported Formats**: text, json, markdown
- **Compatible SDK**: OpenAI

### Capabilities
- Vision (image input)
- Function Calling
- Streaming
- JSON Mode
- System Prompt

### Token Pricing (per 1M tokens)
- **Input Tokens**: 3,251.5 credits ($2.17)
- **Output Tokens**: 26,012 credits ($17.34)
- **Cached Tokens**: 325.15 credits ($0.22)

## Features
- Codex model on the GPT-5.3 base
- Optimized for agentic coding harnesses (Codex-style)
- Served via the Responses API - this gateway routes Codex models to /v1/responses, not chat completions
- 400K context window
- 128K max output tokens
- Cached input tokens for cost savings on repeated context

## Use Cases
- Coding agents and autonomous dev loops
- Complex debugging and architecture-level changes
- Large refactors across a repository
- Code review automation

## API Endpoint

Base URL: https://api.core.today
Endpoint: POST /llm/openai/v1/responses

## Authentication
Header: Authorization: Bearer YOUR_API_KEY

Note: LLM endpoints use OpenAI-compatible format with Authorization Bearer token.

## Input Parameters

### Required
- **model**: string (default: gpt-5.3-codex) - Model identifier
- **input**: string | array - Responses API input: a plain string or an array of input items (messages, tool results)

### Optional
- **instructions**: string - System-level instructions for the run
- **max_output_tokens**: integer - Maximum tokens in the response, including internal reasoning tokens (Responses API field; not max_tokens)
- **stream**: boolean (default: false) - Enable Server-Sent Events streaming

## Examples

### Complex Coding Task
Responses API request using GPT-5.3 Codex
```json
{
  "model": "gpt-5.3-codex",
  "input": "Diagnose why this async job queue occasionally drops tasks under load, then propose and implement a fix:\n\n<code snippet>",
  "max_output_tokens": 6000
}
```

## Response Format
```json
{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Response text here"
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 100,
    "completion_tokens": 50,
    "total_tokens": 150
  }
}
```

## Tips
- Works with the OpenAI SDK - set base_url to https://ai.api.core.today/llm/openai/v1 and use client.responses.create()
- Codex models are served via the Responses API (POST /llm/openai/v1/responses), not chat completions
- Cached input tokens are billed at a 90% discount ($0.175/M) - reuse long repository context across turns

## Documentation
https://platform.openai.com/docs/models