GPT-5.5
OpenAI
OpenAI's flagship as of April 2026 and its first fully retrained base model since GPT-4.5, built agentic-first. GPT-5.5 is OpenAI's strongest agentic coding model and the engine behind Codex, excelling at coding, computer use, knowledge work, and multi-step task completion. Features a 1M-token context window and multimodal (text + vision) input.
Strengths
- OpenAI's strongest agentic coding model and the engine behind Codex
- First fully retrained base model since GPT-4.5, trained agentic-first
- 1M-token context window with multimodal (text + vision) input
- State-of-the-art 82.7% on Terminal-Bench 2.0 (command-line workflows)
- 58.6% on SWE-Bench Pro - more real GitHub issues solved single-pass than prior models
- More token-efficient than GPT-5.4 despite higher per-token price
Caveats
- 2x the per-token price of GPT-5.4 ($5/$30 vs $2.50/$15)
- No audio input (text + vision only)
- Premium tier - GPT-5.4 / 5.4 mini are more cost-effective for simpler tasks
- Vendor
- OpenAI
- Context Window
- 1,000,000 tokens
- Max Output
- 128,000 tokens
- License
- Proprietary
- Release Date
- 2026-04-23
- Training Cutoff
- 2025-12-01
- Modalities
- textimage
| Benchmark | Score | Max |
|---|---|---|
| Terminal-Bench 2.0 | 82.7 | 100.0 |
| SWE-Bench Pro | 58.6 | - |
Capabilities
Vision
Audio
Video
Tool Use
Resources
Pricing
$5 per 1M input tokens, $30 per 1M output tokens (2x GPT-5.4, but more token-efficient). Powers Codex across subscription tiers.