GPT-5.6 Terra
OpenAI·
LLMsmidcloud
Overview
Balanced mid-tier of OpenAI's GPT-5.6 Sol/Terra/Luna family, positioned as the everyday default for interactive and agentic coding. Priced at $2.50/M input, $15/M output tokens. Part of the limited preview from 2026-06-26; generally available 2026-07-09.
Capabilities and innovations
Agentic codingTool use / function callingProgrammatic tool calling (Responses API)Sol / Terra / Luna model family
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| Terminal-Bench 2.x | 77.53% | independent Terminal-Bench 2.1 77.53%, measured by Vals AI with the Terminus 2 harness (accessed 2026-09-23). Vals AI is the reference harness for this field so values compare across vendors; vendor and other third-party figures stay in the notes below. Replaces 82.5 (vendor). |
| Terminal-Bench 4.0 | 21.5% | independent Terminal-Bench 4.0 21.5% (± 3.3, 95% CI), official leaderboard maintained by the Terminal-Bench team (Stanford / Harbor / Laude Institute), max effort, agent harness: Codex (accessed 2026-09-23). 4.0 is its own generation and not comparable with the 2.x or 3.0 fields. |
- GPQA Diamond: independent evaluation pending
- Humanity’s Last Exam (no tools): independent evaluation pending
- SWE-bench Pro: independent evaluation pending
- MMLU-Pro: independent evaluation pending
- Terminal-Bench 3.0: independent evaluation pending
- MMMU-Pro: independent evaluation pending
Links
More from OpenAI
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.