# GPT-5.6 Terra - Core.Today AI API

> The balanced tier of OpenAI's GPT-5.6 family (GA July 2026). Terra offers near-flagship quality at half the price of Sol, with a 1M token context window, 128K max output tokens, and cached inputs at a 90% discount.

- **Provider**: OpenAI
- **Model ID**: gpt-5.6-terra
- **Category**: LLM
- **Credits**: 3 per request
- **Speed**: Fast
- **Quality**: High

## Model Specifications
- **Context Window**: 1M tokens
- **Max Output**: 128K tokens
- **Training Cutoff**: 2026-02
- **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**: 4,645 credits ($3.10)
- **Output Tokens**: 27,870 credits ($18.58)
- **Cached Tokens**: 464 credits ($0.31)

## Features
- Balanced tier of the GPT-5.6 family
- 1M token context window
- 128K max output tokens
- Knowledge cutoff: February 2026
- Cached input pricing (90% discount)
- Adjustable reasoning effort
- Function calling and native vision support

## Use Cases
- Production chat and assistant workloads
- Code generation and review at scale
- Document analysis and summarization
- Agentic workflows with cost constraints
- Repeated context with cached inputs (RAG)

## API Endpoint

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

## Authentication
Header: Authorization: Bearer YOUR_API_KEY

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

## Input Parameters

### Required
- **messages**: array - Array of message objects with role and content
- **model**: string (default: gpt-5.6-terra) - Model identifier

### Optional
- **max_completion_tokens**: integer (default: 4096) - Maximum tokens in response (up to 128000). Note: use max_completion_tokens, not max_tokens
- **reasoning_effort**: string (default: medium) - Reasoning effort level: none, low, medium, high, or xhigh
  Options: none, low, medium, high, xhigh
- **temperature**: float (default: 1.0) - Sampling temperature (0-2)
- **stream**: boolean (default: false) - Enable Server-Sent Events streaming

## Examples

### Production Assistant
Balanced quality and cost for high-volume assistant traffic
```json
{
  "model": "gpt-5.6-terra",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful customer support assistant."
    },
    {
      "role": "user",
      "content": "My webhook deliveries started failing with 403 yesterday. What should I check?"
    }
  ],
  "max_completion_tokens": 2000
}
```

### Code Review
Review a diff with near-flagship reasoning at half the cost
```json
{
  "model": "gpt-5.6-terra",
  "messages": [
    {
      "role": "system",
      "content": "You are a strict code reviewer. Report bugs first."
    },
    {
      "role": "user",
      "content": "<unified diff>"
    }
  ],
  "reasoning_effort": "high",
  "max_completion_tokens": 4000
}
```

## 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
- Best price-performance of the GPT-5.6 family for most production workloads
- Escalate to Sol only for the hardest reasoning tasks; drop to Luna for simple, high-volume calls
- Use cached inputs ($0.25/M, 90% off) for repeated system prompts and RAG context
- reasoning_effort 'high' recovers much of Sol's quality on complex tasks
- Lower temperature (0.2-0.5) for coding and analytical tasks

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