GPT-5.5
GPTFlagship
ThinkingTool UseVisionStructured Output
About this model
GPT-5.5 is OpenAI’s frontier model designed for complex professional workloads, building on GPT-5.4 with stronger reasoning, higher reliability, and improved token efficiency on hard tasks. It features a 1M+ token...
Performance Tier
Flagship
GPT-5.5 is a flagship model from GPT : the most capable in their lineup.
Best-in-class model from this provider. Highest performance across benchmarks, ideal for demanding tasks.
Pricing
This model is included in Elosia plansPremium
Highest cost level. A long conversation can quickly consume your monthly cap.
| Type | per 1M tokens |
|---|---|
| Input (prompt) | $5.00 |
| Output (completion) | $30.00 |
| Cache read | $0.500 |
Capabilities
Context Length1.1M
Max Output Tokens128K
TokenizerGPT
Inputfile, image, text
Outputtext
Release DateApril 24, 2026
Benchmarks
General Intelligence
MMLU
Not reported
GPQA Diamond
93.6%
Mathematics
MATH-500
Not reported
Programming
HumanEval
Not reported
SWE-bench Verified
Not reported
Reasoning
IFEval
Not reported
ARC-AGI-2
85%
Humanity's Last Exam
41.4%
Multimodal
MMMU-Pro
81.2%
Agentic
SWE-bench Pro
58.6%
Terminal-Bench 2.0
82.7%
Recommended Use Cases
CodingAnalysisResearchGeneral ChatData Extraction
Strengths
- 1.05M-token context window with leading long-context retrieval (MRCR 1M ≈ 74)
- Frontier agentic capabilities — leader on Terminal-Bench 2.0 (82.7) and OSWorld-Verified (78.7)
- Top abstract-reasoning score on ARC-AGI-2 (85.0), ahead of Opus 4.7 and Gemini 3.1 Pro
- Natively unified multimodal architecture (text, image, audio, video)
Limitations
- Premium pricing ($5/M input, $30/M output)
- Trails Claude Opus 4.7 on SWE-bench Pro (58.6 vs 64.3) for in-codebase resolution
- OpenAI dropped legacy benchmarks (MMLU, MATH-500, HumanEval) — direct comparison with older models is harder
Resources
This model may use your data for training
