GPT-5.6 Terra: benchmarks, price and rank
GPT-5.6 Terra from OpenAI is 31st of 123 rated AI models on Spezia's leaderboard. It costs $2 in, $12 out per million tokens.
- 31st of 123 rated models
- Rating 1,482
- Rated on 17 benchmarks
- Closed
- Released Jul 9, 2026
- 1.1 million tokens of context
- Knows the world to February 2026
Where it is strongest
Not in the top tenth on any benchmark in use.
Where it falls behind
- 5th of 6 on ARC-AGI-3
- 55th of 68 on SimpleBench
- 16th of 20 on CursorBench
Price: $2 in, $12 out per million tokens.
Its headline results
The benchmarks that count most in the rating, with its place among the models measured the same way, and where each number came from.
- APEX-Agents58.2%
Long tasks written by investment banking analysts, management consultants and corporate lawyers from their own work, done across files and apps in a realistic work setup.
21st of 56. Independent result from Epoch AI.
- Terminal-Bench 4.021.5%
Hard jobs done in a real computer terminal: building and fixing software, setting up systems, and science and engineering problems.
Independent result from LLM Stats.
- BrowseComp87.5%
Questions whose answers are hidden deep on the web and take many searches and pages to track down.
7th of 56. The maker's own number, from OpenAI, via LLM Stats.
- SWE-Bench Pro63.4%
Long changes to real codebases in several languages, often across many files, including code from private and copyleft projects that models are unlikely to have seen.
11th of 50. The maker's own number, from OpenAI, via LLM Stats.
- ARC-AGI-30.8%
Interactive puzzle games the model has never seen, with no instructions. It has to explore, work out the goal and the rules, and win.
5th of 6. The maker's own number, from OpenAI, via LLM Stats.
- Vending-Bench 2$7,343
The model runs a simulated vending machine business for a year on its own: finding suppliers, ordering stock, setting prices and paying the daily fees.
19th of 60. Independent result from Epoch AI.
- Toolathlon53.1%
Long tasks across many real apps and their tools, such as email, calendars, spreadsheets, code repositories and online stores.
16th of 37. The maker's own number, from OpenAI, via LLM Stats.
- CursorBench41.3%
Ambiguous coding tasks that span many files, taken from real sessions in the Cursor code editor.
16th of 20. Independent result from Epoch AI.
- DeepSWE 1.170%
Software engineering tasks run in the same simple agent setup for every model.
Independent result from LLM Stats.
- SimpleBench48.9%
Multiple-choice puzzles about everyday situations, space, time and people, written to trip up models that pattern-match instead of think. Ordinary adults do well on them.
55th of 68. Independent result from Epoch AI.
14 more results in its full profile.
Every result with its source, what it costs per task, what jumped with its release, side by side with any other model, and answers to your questions about when to use it.
Models ranked near it
How the rating works, and every benchmark it uses: the model leaderboard.