OpenAI Releases GPT-5.6 Terra: Mid-Tier Model at $2.50 Input/$15 Output per 1M Tokens

OpenAI has released GPT-5.6 Terra, a balanced model in its GPT-5.6 series priced at $2.50 per million input tokens and $15 per million output tokens. The model features a 1 million token context window and a February 2026 knowledge cutoff date.

Positioning and Pricing

GPT-5.6 Terra sits between OpenAI's flagship Sol tier and cost-efficient Luna tier within the GPT-5.6 family. According to OpenAI, Terra offers "strong performance at roughly half the cost of Sol," targeting users who need to balance capability with cost efficiency.

The model is released with a July 9, 2026 availability date, making this a future release announcement rather than an immediate launch.

Technical Specifications

• Context Window: 1,000,000 tokens
• Input Pricing: $2.50 per 1M tokens
• Output Pricing: $15 per 1M tokens
• Knowledge Cutoff: February 2026
• Modalities: Text input and output

Target Use Cases

OpenAI positions Terra for "everyday coding, reasoning, and agentic tasks" where users require capable performance without flagship-tier pricing. The model is designed for production workloads that need consistent quality at scale.

The 1M token context window enables handling of large codebases, extended conversations, and document analysis tasks without chunking or summarization.

Provider Availability

The model is hosted through OpenRouter, which forwards requests directly to OpenAI without routing decisions. OpenRouter reports that actual customer costs can run 60-80% lower than list prices when accounting for prompt caching with repeated context.

Benchmark scores, parameter count, and detailed capability comparisons with Sol and Luna tiers have not been disclosed.

What This Means

GPT-5.6 Terra represents OpenAI's continued strategy of offering tiered model options within a single generation, allowing developers to optimize for their specific cost-performance requirements. The 1M context window at $2.50/$15 pricing positions it as a direct competitor to Claude 3.5 Sonnet and other extended-context models in the market. The February 2026 knowledge cutoff suggests this model will maintain current information for several months post-release, though the July 2026 availability date indicates this is a planned future launch rather than an immediate deployment.