GPT-5.6 Terra API Pricing: OpenAI's Mid-Tier Model at $2/M Input

GPT-5.6 Terra replaces GPT-5 mini and o4-mini with 1.05M context at $2/M input, $12/M output — the sweet spot between cost and capability in OpenAI's lineup.

Updated Aug 25, 2026 · 147 models tracked across 22 providers

TL;DR

Availability: Terra is also generally available through Amazon Bedrock in AWS GovCloud (US-East and US-West), and as a model inside Kiro. The Bedrock GovCloud route is a separate provider surface with its own access and billing; do not treat the direct OpenAI API price below as a quoted Bedrock price.

Migration Notice: GPT-5 mini shuts down Dec 11, 2026 and o4-mini shuts down Oct 23, 2026. GPT-5.6 Terra is the recommended replacement for both — offering 3.8x more context than GPT-5 mini and 5.25x more than o4-mini at a comparable price point.

GPT-5.6 Terra Pricing Breakdown

At $2.00 per million input tokens and $12.00 per million output tokens, Terra sits between Sol ($4/$20) and Luna ($0.20/$1.20). Here's how the costs scale:

Monthly Volume Input Cost Output Cost Total Cost
1M tokens$2.00$12.00$14.00
10M tokens$20.00$120.00$140.00
100M tokens$200.00$1,200.00$1,400.00
1B tokens$2,000.00$12,000.00$14,000.00

At 100M tokens/month, GPT-5.6 Terra costs $1,400. The same volume on GPT-5.6 Sol would cost $2,750, and on GPT-5.6 Luna it would cost $140.

GPT-5.6 Family: Sol vs Terra vs Luna

The GPT-5.6 family shares the same 1.05M context window but differs in capability and cost:

Model Input $/M Output $/M Context Best For
GPT-5.6 Sol $4.00 $20.00 1.05M Flagship reasoning, research, nuanced writing
GPT-5.6 Terra $2.00 $12.00 1.05M Complex analysis, coding, agentic workflows
GPT-5.6 Luna $0.20 $1.20 1.05M Volume tasks, RAG, bulk processing

Terra is 50% cheaper than Sol on both input and output, while offering the same context window. For most production workloads that need strong reasoning but not absolute top-tier capability, Terra is the best value in the family.

How GPT-5.6 Terra Compares to Other Mid-Tier Models

Model Input $/M Output $/M Context Vision Provider
Claude Sonnet 5 $2.00 $10.00 1M ✅ Anthropic
GPT-5.6 Terra $2.00 $12.00 1.05M ✅ OpenAI
GPT-5.4 $2.50 $15.00 272K ✅ OpenAI
Grok 4.5 $2.00 $6.00 1M ✅ xAI
Mistral Medium 3.5 $1.50 $7.50 128K ✅ Mistral
DeepSeek V4 Pro $0.66 $1.98 1M ❌ DeepSeek

Terra matches Claude Sonnet 5 on input pricing and offers a slightly larger context window (1.05M vs 1M). Grok 4.5 is cheaper on output, but Terra has stronger ecosystem integration with OpenAI's function calling, structured outputs, and tooling.

Migrating to GPT-5.6 Terra

If you're currently using GPT-5 mini or o4-mini, here's what changes when you migrate:

FeatureGPT-5 minio4-miniGPT-5.6 Terra
Input price$0.25/M$1.10/M$2.00/M
Output price$2.00/M$4.40/M$12.00/M
Context window272K200K1.05M
Shutdown dateDec 11, 2026Oct 23, 2026Active
ReasoningStandardChain-of-thoughtAdaptive thinking
Vision✅✅✅

Key trade-off: Terra is more expensive per token than both GPT-5 mini and o4-mini, but offers 3.8x–5.25x more context and improved reasoning with adaptive thinking. If you were using o4-mini for chain-of-thought reasoning, Terra's adaptive thinking provides similar capabilities.

When to Use GPT-5.6 Terra

Complex Coding

Multi-file refactoring, architecture decisions, and code review. Terra's reasoning handles complex dependencies better than Luna.

Agentic Workflows

Multi-step tool-calling agents that need strong reasoning at each step. Terra balances cost and capability for production agents.

Data Analysis

Complex data interpretation, statistical analysis, and report generation. Terra handles nuanced analysis that Luna struggles with.

Content Generation

Technical writing, documentation, and marketing copy that needs accuracy and nuance. Better quality than Luna at 10x the cost.

Research Tasks

Literature review, fact-checking, and synthesis across large document sets. 1.05M context fits most research corpora.

Prototyping

Building and testing AI features before optimizing for cost. Start with Terra, then move stable workloads to Luna for savings.

Real-World Cost: 10,000 Agentic Coding Tasks/Month

Suppose you're running an AI coding assistant that handles 10,000 tasks per month, averaging 8,000 input tokens and 2,000 output tokens per task:

ModelInput CostOutput CostMonthly Total
GPT-5.6 Luna$16.00$24.00$40.00
GPT-5.6 Terra$160.00$240.00$400.00
Claude Sonnet 5$160.00$200.00$360.00
GPT-5.6 Sol$320.00$400.00$720.00

Terra handles the workload for $400/month — 45% cheaper than Sol. Claude Sonnet 5 is slightly cheaper on output, but Terra's larger context window (1.05M vs 1M) may reduce the number of calls needed for complex tasks.

Compare 147 Models Side by Side

GPT-5.6 Terra is one of 147 models tracked on APIpulse. Compare pricing, context windows, and features across 22 providers.

Frequently Asked Questions

How much does GPT-5.6 Terra cost?
GPT-5.6 Terra costs $2.00 per million input tokens and $12.00 per million output tokens. This positions it between GPT-5.6 Sol ($4/$20) and GPT-5.6 Luna ($0.20/$1.20) in the GPT-5.6 family.
What models does GPT-5.6 Terra replace?
GPT-5.6 Terra replaces both GPT-5 mini (shutting down Dec 11, 2026) and o4-mini (shutting down Oct 23, 2026). Terra offers the same 1.05M context as Sol at 50% lower cost, making it the natural upgrade path for mid-tier workloads.
What is the context window of GPT-5.6 Terra?
GPT-5.6 Terra has a 1.05 million token context window — the same as GPT-5.6 Sol and GPT-5.6 Luna. This is significantly larger than GPT-5 mini's 272K context and o4-mini's 200K context.
Is GPT-5.6 Terra cheaper than Claude Sonnet 5?
Yes. GPT-5.6 Terra ($2/$12) is the same price as Claude Sonnet 5 ($2/$10) on input but slightly more expensive on output ($12 vs $10). Both are mid-tier models with 1M context windows. Terra is better for OpenAI ecosystem integration; Sonnet 5 is better for output-heavy workloads.
When should I use GPT-5.6 Terra instead of Sol or Luna?
Use Terra for complex analysis, coding, and agentic workflows that need strong reasoning but don't require Sol's maximum capability. Terra is 50% cheaper than Sol on both input and output. Use Luna ($0.20/$1.20) for simple high-volume tasks; use Sol ($4/$20) when you need the absolute best reasoning.
Does GPT-5.6 Terra support vision?
Yes. GPT-5.6 Terra supports image input alongside text, like all GPT-5.6 models. This makes it suitable for multimodal tasks like document analysis, image understanding, and visual Q&A at a mid-tier price point.

Related Pages