Model
GPT-5.5
Default frontier GPT for coding, computer use, research, and knowledge work
20 providersreleased Apr 23, 2026knowledge cutoff 2025-12-01closed weights
$4.55
best input /M · Poe
$27.27
best output /M
1.05M
context window
128K
max output
Compare providers (sorted by listed input price)
| Provider | Input /M | Output /M | Cache read /M | Cache write /M | Context | Capabilities | Status | Updated |
|---|---|---|---|---|---|---|---|---|
| Poe openai/gpt-5.5 | $4.55 | $27.27 | $0.455 | — | 400K | reasoningtoolsstructuredvision | Apr 8, 2026 | |
| HPC-AI openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| OrcaRouter openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| NanoGPT openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Impossibl openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| OpenRouter openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| CrossModel openai/gpt-5.5 | $5 | $30 | $0.5 | $5 | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| OpenAI gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| LLM Gateway openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| NEAR AI Cloud openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Merge Gateway openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Cloudflare AI Gateway openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| FastRouter openai/gpt-5.5 | $5 | $30 | — | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| ZenMux openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Perplexity Agent openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsno structuredvision | Apr 23, 2026 | |
| GMI Cloud openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Ofox openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Vercel AI Gateway openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Eden AI openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 | |
| Kilo Gateway openai/gpt-5.5 | $5 | $30 | $0.5 | — | 1.05M | reasoningtoolsstructuredvision | Apr 23, 2026 |
Price badge
https://models.sutraworks.ai/badge/openai/gpt-5.5.svgLive SVG, regenerated on every hourly sync — paste it into a README to always show the current cheapest listed price. Updates when prices change; no build step needed on your side.
In the news
All news →- GPT-5.6: Frontier intelligence that scales with your ambition
openai.com· 3d ago - Previewing GPT-5.6 Sol: a next-generation model | OpenAIopenai.com· 3d ago
Benchmarks
- SWE-Bench Pro(resolve rate)58.6
- Terminal-Bench(success rate)78.2
- SWE-Atlas Codebase QnA(score)45.43
- SWE-Atlas Refactoring(score)44.79
- SWE-Atlas Test Writing(score)42.59
- Artificial Analysis Coding Agent Index(average pass@1)65.3
- SWE-Atlas Codebase QnA(pass@1)80.8
- SWE-Bench Pro(pass@1)30.9
- Terminal-Bench(pass@1)84.1
- Artificial Analysis Coding Agent Index(average pass@1)60.4
- SWE-Atlas Codebase QnA(pass@1)79.1
- SWE-Bench Pro(pass@1)26.2
- Terminal-Bench(pass@1)75.8
- Artificial Analysis Coding Agent Index(average pass@1)57.8
- SWE-Atlas Codebase QnA(pass@1)75
- SWE-Bench Pro(pass@1)24.9
- Terminal-Bench(pass@1)73.4
- Terminal-Bench(success rate)82.7
- GPQA Diamond(accuracy)93.6
- Humanity's Last Exam(accuracy)41.4
- Humanity's Last Exam(accuracy)52.2
- OSWorld-Verified(success rate)78.7
- BrowseComp(accuracy)84.4
- MMMU Pro(accuracy)81.2
- ARC-AGI-2(accuracy)85
- FrontierMath(accuracy)51.7
- FrontierMath(accuracy)35.4
- GDPval(wins or ties)84.9
- MCP Atlas(success rate)75.3
- Toolathlon(success rate)55.6
- τ²-Bench Telecom(success rate)98
Prices are per million tokens (USD). “—” means the provider does not publicly list a price for this model. Data from models.dev; verify with the provider before purchasing.