GPT-5.5
openai-gpt-5-5
fields
| nameREQ | GPT-5.5 | openai.com | 2026-04-30 |
| vendorREQ | OpenAI | openai.com | 2026-04-30 |
| released_at | 2026-04-23 | openai.com | 2026-04-30 |
| modality | text, code, vision | openai.com | 2026-04-30 |
| access | closed | openai.com | 2026-04-30 |
| context_window | 1000000 | openai.com | 2026-04-30 |
| parameters | |||
| benchmark_score | 88.72 revisions | www.buildfastwithai.com | 2026-05-04 |
| benchmark_name | SWE-Bench Verified2 revisions | www.buildfastwithai.com | 2026-05-04 |
| announcement_url | https://openai.com/index/introducing-gpt-5-5/ | openai.com | 2026-04-30 |
history · 9 fields · 11 revisions
name1 revision
GPT-5.5
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
vendor1 revision
OpenAI
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
released_at1 revision
2026-04-23
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
modality1 revision
text, code, vision
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
access1 revision
closed
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
context_window1 revision
1000000
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
benchmark_score2 revisions
88.7
current
GPT-5.5 leads on SWE-bench Verified at 88.7% (vs Claude Opus 4.7's 87.6%). Also scored 82.7% on Terminal-Bench 2.0.
82.7
superseded
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
benchmark_name2 revisions
SWE-Bench Verified
current
GPT-5.5 SWE-bench Verified 88.7%. Terminal-Bench 2.0 82.7%.
Terminal-Bench 2.0
superseded
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.
announcement_url1 revision
https://openai.com/index/introducing-gpt-5-5/
current
We're releasing GPT-5.5, our smartest and most intuitive to use model yet. Terminal-Bench 2.0: 82.7%. Context window: 1M tokens.