OpenAI Releases GPT‑5.5 Instant, Says It Is Smarter, Clearer, Personalized
OpenAI claims that the model can reduce hallucination in law, medicine, and finance while maintaining low latency.
News
- US to Test Frontier AI Models Before Release in Safety Push
- OpenAI Releases GPT‑5.5 Instant, Says It Is Smarter, Clearer, Personalized
- Railways Begins AI-Based Workforce Verification Rollout
- PNB Steps Up Cybersecurity Spending as AI Risks Rise
- Anthropic Forms AI Services Firm With Wall Street Backers
- Pixxel, Sarvam Plan India’s First Orbital Data Center Satellite
Following the launch of the GPT-5.5 model late last month, the ChatGPT maker, on Tuesday released yet another foundation model, GPT-5.5 Instant. OpenAI said this model will replace GPT-5.3 Instant as the default ChatGPT model and that it is smarter, cleaner, and more personalized.
In a blog post, the company claimed that this model can reduce hallucination in sensitive areas such as law, medicine, and finance while maintaining the low latency of its predecessor.
While releasing GPT 5.5, the company said that this model had improved coding and knowledge work capabilities.
The GPT 5.3 model achieved a score of 81.2 on the AIME 2025 math test, compared to 65.4 for the older model. OpenAI said that the model outperformed its predecessor on the MMMU-Pro multimodal reasoning benchmark, with a score of 76 vs. 69.2..
A math reasoning benchmark is built from American Invitational Mathematics Examination–style questions while MMMU‑Pro is a multimodal benchmark that checks how well a model can reason across text and images in many specialized domains like science, engineering, etc.
The release said that the new model offers improved context management and a search tool to refer back to past conversations, files, and Gmail to provide personalized responses. OpenAI said this feature will be offered to Plus and Pro users of this model on the web and that it will also be rolled out to mobile users soon.
OpenAI said that it plans to extend access to this feature to Free, Go Business, and enterprise users in the coming weeks.
With this latest launch, ChatGPT will display the sources of information to help users understand where the information comes from. It added that users can delete outdated sources or correct the sources if the answers were wrong. The company said that if users share a chat with someone, they won’t be able to see the memory sources.
ChatGPT maker said that the GPT 5.5 model will be available via the API as chat-latest for developers, with 5.3 available as an option for paid users for 3 months.
OpenAI released GPT-5.5 in the last week of April and called it “the smartest and most intuitive” model yet, signaling its further push into coding, knowledge work, and agentic AI.
The company announced that GPT-5.5 is designed for complex tasks that require reasoning across contexts and sustained action over time, including writing and debugging code, online research, data analysis, document creation, spreadsheet work, and software use.
OpenAI said the model can handle messy, multi-part tasks by planning, using tools, checking its work, and moving across applications with fewer step-by-step instructions from users.
It may be recalled that when OpenAI withdrew its GPT-4o model, there was significant backlash from users who felt the model had a “personality.”
GPT-4 often affirmed users’ choices, which made them feel a connection to that particular model. Following the withdrawal of this model, users signed petitions to stop OpenAI from retiring it, calling it their “best friend” or “a mirror.”
Despite the outcry, GPT-4o was deprecated in February 2026.
