OpenAI API launches GPT-5.4, adds 10% data residency surcharge
What changed
Flagship model lineup restructured: GPT-5.2 and GPT-5.2 pro replaced by GPT-5.4 [GPT-5.2 input $1.75/1M, output $14/1M; GPT-5.2 pro input $21/1M, output $168/1M; GPT-5.4 input $2.50/1M, output $15/1M; GPT-5.2 pro eliminated entirely]
gpt-realtime-1.5 model added to Realtime API with same pricing as gpt-realtime [Text: $4/1M input, $0.40/1M cached, $16/1M output; Audio: $32/1M input, $0.40/1M cached, $64/1M output; Image: $5/1M input, $0.50/1M cached]
Pricing caveat expanded: standard rates now specified for context lengths under 270K; 10% surcharge added for data residency and regional processing on GPT 5.4 models
Before and after
Mar 2, 2026 vs Mar 9, 2026Flagship model lineup restructured: GPT-5.2 and GPT-5.2 pro replaced by GPT-5.4 [GPT-5.2 input $1.75/1M, output $14/1M; GPT-5.2 pro input $21/1M, output $168/1M; GPT-5.4 input $2.50/1M, output $15/1M; GPT-5.2 pro eliminated entirely]
Sign in to view the full comparison
See the before/after screenshots and every marked-up change.
More from OpenAI
Full history →- Week 41, 2026·improvedOpenai API halves GPT-6.1 Sol cached input price to $0.10 per 1M tokens
- Week 36, 2026·packagingOpenai API Business: new $100/mo Premium seat tier + 200-employee cap
- Week 20, 2026·increaseOpenai API raises GPT-Realtime-2 text output 50%, adds translate
- Week 18, 2026·decreaseOpenai API cuts image output pricing, launches GPT-image-2
- Week 14, 2026·decreaseOpenai API slashes web search tool price 60% to $10 per 1k calls
