Loading model details…
OpenAI's next-generation general-purpose model balancing intelligence and cost, featuring long-context reasoning, multimodal understanding, and full agent tooling for enterprise AI, coding, automation, and professional knowledge work.
OpenAI
| Item | Specification |
|---|---|
| Input | Text, Image |
| Output | Text |
| Context | 1,050,000 tokens |
| Max Output | 128,000 tokens |
| Vision | ✓ Supported |
| Function Calling | ✓ Supported |
GPT-5.6 Terra is the balanced model in OpenAI's GPT-5.6 family, released on July 9, 2026 for production workloads that value both intelligence and cost. It roughly corresponds to the mini tier in earlier GPT-5 families.
At a lower price than the flagship tier, the model retains a 1,050,000-token context window, a 128,000-token maximum output, six reasoning-effort levels, image understanding, function calling, and Structured Outputs. As part of the GPT-5.6 family, it also supports persisted reasoning, Programmatic Tool Calling, explicit prompt caching, multi-agent orchestration in beta, and Responses API tools for retrieval, code execution, computer interaction, MCP, and Skills.
none, low, medium, high, xhigh, and max to control quality, latency, and consumption by task difficulty.original image detail support interface screenshots, charts, scanned documents, and high-resolution material.| Feature | Benefit |
|---|---|
| Intelligence-Cost Balance | Current rates of $2 per million input tokens and $12 per million output tokens provide strong capability for production deployments with quality and budget requirements. |
| Six Reasoning Levels | Simple tasks can use less reasoning for speed, while difficult tasks can receive more reasoning effort for higher quality. |
| 1.05M-Token Context | Processes large codebases and long source collections in one request, reducing information loss and engineering overhead from chunking. |
| Persisted Reasoning | Reuses reasoning state across turns and consecutive tool calls, reducing repeated computation and improving workflow continuity. |
| Programmatic Tool Calling | Allows tools to be invoked during code execution, shortening round trips in complex agent tasks. |
| Broad Hosted Tool Ecosystem | Web Search, File Search, Hosted Shell, Computer Use, MCP, and Skills cover a complete retrieval-to-action workflow. |
Tiered by input prompt tokens (incl. cache): once over the threshold, the whole request is billed at the higher tier.
| Token Type | Short context ≤272K | Long context >272K |
|---|---|---|
| Input | $1.5$2-25% | $3$4-25% |
| Output | $9$12-25% | $13.5$18-25% |
| Cached input | $0.15$0.2-25% | $0.3$0.4-25% |
| Cache write | $1.875$2.5-25% | $3.75$5-25% |