Strengths
- OpenAI says it nearly matches the pricier GPT-6.2 Astra on the FrontierCode 2.0 coding test (43.8% vs 44.2%) at roughly a fifth of the cost per task.
- Stronger than GPT-6.1 Sol at the same price. It scores 57 on the independent Artificial Analysis index v5.0.1 against 55 for its predecessor, at $1.01 per test task instead of $1.27.
- On OpenAI's internal GDPval v2 test of long real-world projects it improved from 41.5% for GPT-6.1 Sol to 46.9%.
- Writes faster than its predecessor: about 77 tokens per second according to Artificial Analysis, against about 57 for GPT-6.1 Sol.
- Cached input costs only $0.10 per million tokens, 95% less than regular input. That saves money for agents that keep working with the same context.
Weaknesses
- Not in regular ChatGPT chat, only in ChatGPT Work and Codex. Chat keeps the older models.
- For the hardest scientific work, GPT-6.2 Astra is stronger. OpenAI reports 68.0% for Sol on Terminal-Bench-Science 0.1 against 77.1% for Astra.
- At the highest thinking level it takes a while to start: about 76 seconds to the first word of the answer, according to Artificial Analysis.
- Like Astra, it has stricter cybersecurity filters that also hold back some legitimate security testing, OpenAI says.
Best for
- the default model for business apps and agents via the API
- coding at a reasonable cost
- everyday work with documents and spreadsheets in ChatGPT Work
- automated multi-step workflows
Not ideal for
- high volumes of simple tasks, where GPT-6.1 Terra and GPT-6 Luna are cheaper
- the most demanding scientific research
How to use it in Europe
In ChatGPT it comes with Plus, Pro, Business, Enterprise and Edu, but only in ChatGPT Work and Codex, not in regular chat. The faster Ultrafast mode is on Pro 500 and now also on Pro 200. Codex CLI needs updating to the latest version. Developers use it through the OpenAI API, which is available in Slovakia and Czechia. It supports EU data processing, which costs 10% extra in the API.
Price compared with similar models
Benchmark results
| Benchmark | Score | Source |
|---|---|---|
| Artificial Analysis Intelligence Index v5.0.1 (max effort) | 57 | artificialanalysis.ai ↗, November 2026 |
| FrontierCode 2.0, max effort | 43.8% | openai.com ↗ (vendor-reported), November 2026 |
| GDPval v2, max effort | 46.9% | openai.com ↗ (vendor-reported), November 2026 |
| Terminal-Bench-Science 0.1, max effort | 68.0% | openai.com ↗ (vendor-reported), November 2026 |
| LifeVal, max effort | 83.8% | openai.com ↗ (vendor-reported), November 2026 |
Benchmarks cover only a narrow slice of ability and vendors often report them at their best settings. Treat them as a rough guide.
More models from this vendor
- GPT-6.2 Astra: OpenAI's most capable model for demanding professional work, coding and science. A fifth cheaper than GPT-6 Astra.
- GPT-6.1 Terra: OpenAI's fast, cheaper model for working with images, operating a computer and everyday office work.
- GPT-6 Luna: OpenAI's cheapest model, made for high volumes of simple tasks, sorting text and low-cost chatbots.
- gpt-oss-120b: OpenAI's open model that a company can run on its own server when data must not leave its infrastructure.
- gpt-oss-20b: OpenAI's smaller open model that runs on a computer with 16 GB of memory, even without internet access.
Similar models from other vendors
- Claude Sonnet 5.5 (Anthropic): Claude's balanced model for everyday work and coding. It is faster than Opus 5.5 and costs half as much.
- Command A+ (Cohere): An open business model from Canada's Cohere that officially supports Slovak and Czech and runs on your own server.
- Gemini 4 Hydrogen (Google): Google's faster model for everyday coding and agents, out since 12 November 2026. It has a free API tier and comes with Google AI Plus and up.
- Muse Spark 1.3 (Meta): Meta's balanced model with a 1-million-token window and a fair price. The Muse Spark line also powers the Meta AI assistant.
- Muse Spark Open (Meta): Meta's large open model that a company can run on its own servers, so data never leaves its infrastructure.
- Llama 4 Maverick (Meta): The larger model in the newest Llama generation, from April 2025. Cheap through providers, but its license excludes people and companies in the EU.