Introducing GPT-5.5
OpenAI Blog · 2026-04-23
(No summary yet for this item — extraction summaries are still backfilling.)
Appears in
Extraction
Topics: gpt-5-5large-language-model-releaseagentic-codingmodel-benchmarksai-safety
Claims
- GPT-5.5 achieves 82.7% on Terminal-Bench 2.0, 73.1% on Expert-SWE (internal), and 58.6% on SWE-Bench Pro, outperforming GPT-5.4 on all three while using fewer tokens.
- GPT-5.5 matches GPT-5.4 per-token latency in real-world serving despite operating at a higher level of intelligence.
- GPT-5.5, using a custom harness, helped discover a new proof about off-diagonal Ramsey numbers that was subsequently verified in Lean.
- OpenAI classifies GPT-5.5's biological/chemical and cybersecurity capabilities as 'High' under its Preparedness Framework.
- GPT-5.5 co-designed inference optimizations on NVIDIA GB200/GB300 hardware that increased token generation speeds by over 20%.
- GPT-5.5 Pro is priced at $30 per 1M input tokens and $180 per 1M output tokens in the API; standard GPT-5.5 is $5 input / $30 output per 1M tokens.
Key quotes
The first coding model I've used that has serious conceptual clarity. — Dan Shipper, Founder and CEO of Every
GPT-5.5 is noticeably smarter and more persistent than GPT-5.4, with stronger coding performance and more reliable tool use. It stays on task for significantly longer without stopping early, which matters most for the complex, long-running work our users delegate to Cursor.
Losing access to GPT-5.5 feels like I've had a limb amputated. — Engineer at NVIDIA with early access