GPT-5.5
OpenAI·
LLMsflagshipcloud
Overview
Agentic-focused flagship with 1M context window (400K in Codex). Thinking and Pro variants available. Codename 'Spud'.
Capabilities and innovations
1M context windowThinking modePro variant ($30/$180 per M tokens)400K context in CodexAgentic workflowsAPI pricing: $5/$30 per M tokens (base)1M Context Window (largest OpenAI flagship)Codex-Optimized 400K ContextThinking + Pro Variant Lineup
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Pro | 58.6% | vendor |
| MMLU-Pro | 88.14% | unsourced |
| Terminal-Bench 2.x | 83.4% | vendor Terminal-Bench 2.1: 83,4% (OpenAI self-report). |
- GPQA Diamond: not reported by the vendor
- Humanity’s Last Exam (no tools): not reported by the vendor
- MMMU-Pro: not reported by the vendor
Reliability
- Hallucination rate (Vectara HHEM)
- 9.3% (lower is better)independent
Links
More from OpenAI
Data curated by AI Model Timeline. See the methodology for admission criteria, benchmark eras and source priority.