GPT-5.6 Terra
GPT-5.6 Terra is an AI model from OpenAI, released on 2026-06-26.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Model basics
Open source & experience
Official resources
API details
| Type | Condition | Input | Output |
|---|---|---|---|
| Text | - | $2.50/ 1M | $15.00/ 1M |
| Type | Condition | Input | Output |
|---|---|---|---|
| Text | context_tokens > 272000 | $5.00/ 1M | $22.50/ 1M |
| Type | TTL | Write | Read |
|---|---|---|---|
| Text | 30m | $3.13/ 1M | $0.250/ 1M |
Benchmark Results
GPT-5.6 Terra currently shows benchmark results led by ARC-AGI (2 / 68, score 96.50), ARC-AGI-2 (7 / 62, score 83.90), DeepSWE (3 / 20, score 69.60). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.
General Knowledge
5 evaluationsCoding and Software Engineer
2 evaluationsCompare with other models
No curated comparisons for this model yet.
Want a custom combination? Open the compare tool
Publisher
Model Overview
GPT-5.6 Terra
GPT-5.6 Terra is OpenAI's balanced mid-tier model in the GPT-5.6 family, first previewed on June 26, 2026 and made generally available on July 9, 2026. It targets high-concurrency production workloads — customer support, internal tooling, document analysis — delivering roughly GPT-5.5-level capability at about half the price. The family also includes Sol (the flagship, for the hardest coding/research/security tasks) and Luna (the most cost-efficient tier).
On Artificial Analysis's Intelligence Index (max reasoning), Terra scores 55 versus Sol's 59, at roughly half the cost per task (~$0.55 vs ~$1.04). It reaches 87.40 on Terminal-Bench 2.1, 69.60 on DeepSWE, 77.4 on the AA Coding Agent Index, 50.4 on Agents' Last Exam, and 65.14% accuracy on the Vals Index. Independent testing by CodeRabbit found Terra completes 40.7% of long-horizon coding tasks (versus Sol's 63.7%) while producing notably more verbose output per task — best suited to well-scoped, cost-sensitive coding work rather than long or highly complex jobs.
Terra shares OpenAI's GPT-5.6 system card with Sol and Luna, including "High" Preparedness Framework ratings for cybersecurity and biological/chemical risk; see the GPT-5.6 Sol page for the detailed safety-evaluation breakdown.
DataLearner on WeChat
Follow DataLearner on WeChat for AI model updates and research notes.
