Models
20 · 2024–2026
GPT Realtime TranslateAudio16K
GPT Realtime WhisperAudio16K
GPT-4oMultimodal · LLM · Tool-using model +2128K
GPT-4.1LLM · Multimodal · Tool-using model1M
GPT-5 ThinkingReasoning model · Multimodal · LLM400K
GPT-5.1LLM · Multimodal · Reasoning model +1400K tokenów wejścia; 128K tokenów wyjścia
GPT-5.2LLM · Multimodal · Reasoning model +1Nieujawnione publicznie w komunikacie OpenAI dla GPT-5.2 (rodzina GPT-5 standardowo operuje w zakresie 400K tokenów wejścia)- GPT-5.3-Codex-SparkLLM · Tool-using model128K
GPT-5.4 ProLLM · Multimodal · Reasoning model +21.05M
GPT-5.4 Thinking1,050,000- GPT-5.4 miniLLM · Multimodal400,000 tokens (max output 128,000)
ChatGPT Images 2.0Image generation · Reasoning model · Multimodal
GPT-5.5LLM · Multimodal · Reasoning model +11M
GPT-5.5 ProLLM · Multimodal · Reasoning model +11M
GPT-5.5 InstantLLM · Multimodal
GPT Realtime 2Audio · Multimodal · Reasoning model128K- GPT-Live-1Audio · Multimodal
- GPT-Live-1 miniAudio · Multimodal
- GPT-5.6 SolLLM · Reasoning model400K
GPT-RedLLM · Reasoning modelnieujawnione