Claims and evidence
- Released Primary source[1]Availability [4]Mar 3 (2026)
- Deprecated Primary source[5]2026-05-08: gpt-5.2-chat-latest and gpt-5.3-chat-latest model snapshots
- Retired Primary source[5]2026-05-08: gpt-5.2-chat-latest and gpt-5.3-chat-latest model snapshots
- Status RetiredIts only API form, gpt-5.3-chat-latest, was shut down on 2026-08-10. In ChatGPT it was replaced as the default by GPT-5.5 Instant on 2026-05-05 and stayed available to paid users for three months before retirement; no dated ChatGPT retirement entry was found. Primary source[5]2026-05-08: gpt-5.2-chat-latest and gpt-5.3-chat-latest model snapshots [6]May 5, 2026: GPT-5.5 Instant in ChatGPT
- Replaced by GPT-5.5 InstantGPT-5.5 Instant replaced it as the ChatGPT default and in the API as chat-latest. The deprecations page names gpt-5.6-sol as the replacement for gpt-5.3-chat-latest. Primary source[7]Availability [6]May 5, 2026: GPT-5.5 Instant in ChatGPT
- Successor of GPT-5.2The post compares GPT-5.3 Instant with GPT-5.2 Instant throughout and moves GPT-5.2 Instant to the legacy models; GPT-5.2 Instant is part of the GPT-5.2 record. Primary source[1] [8]Introduction
- Change · Safety OpenAI says it gives fewer unnecessary refusals and fewer defensive or moralizing caveats than GPT-5.2 Instant.Compared with GPT-5.2 Primary source[1]
- Change · Tool use OpenAI says it balances web search results better with its own knowledge, for example to put recent news in context.Compared with GPT-5.2 Primary source[1]
- Change · Other OpenAI reports fewer hallucinations in its internal evaluations, with or without web access.Compared with GPT-5.2 Primary source[1]
- Change · Other OpenAI describes a more focused conversational style with fewer unwarranted assumptions about user intent or emotions.Compared with GPT-5.2 Primary source[1]
- Input text, image Primary source[2]Model details
- Output text Primary source[2]Model details
- Feature Function calling Primary source[2]Supported features
- Context window 128,000 tokensThe docs also list a maximum of 272,000 input tokens. Primary source[2]Model details
- Parameters Not disclosedNot disclosed by OpenAI. No claim madeNo source
- Access consumer app, API Primary source[1]Availability
Lineage
Predecessors
- GPT-5.2 · 11 December 2025
Successors
- GPT-5.5 Instant · 5 May 2026
Based on
Not derived from another model.
Variants and derived
None recorded.
Siblings
None recorded.
All ancestors
- GPT · 11 June 2018
- GPT-2 · 14 February 2019
- GPT-3 · 11 June 2020
- GPT-3.5 · 15 March 2022
- GPT-4 · 14 March 2023
- GPT-4 Turbo · 6 November 2023
- GPT-4o · 13 May 2024
- o1-mini · 12 September 2024
- o1-preview · 12 September 2024
- o1 · 5 December 2024
- o3-mini · 31 January 2025
- GPT-4.1 · 14 April 2025
- o3 · 16 April 2025
- o4-mini · 16 April 2025
- o3-pro · 10 June 2025
- GPT-5 · 7 August 2025
- GPT-5.1 · 12 November 2025
- GPT-5.2 · 11 December 2025
All descendants
- GPT-5.5 Instant · 5 May 2026
Variants
No variants recorded in this record.
Related AI Radar coverage
AI Radar coverage starts in June 2026; no coverage linked yet.