OpenAI Launches ChatGPT-5 – Here’s What’s Changed

OpenAI on Thursday announced the launch of ChatGPT-5 with CEO Sam Altman saying that GPT-5 is a “major upgrade” over GPT-4 and a “significant step” in the company’s path to artificial general intelligence (AGI).

Altman said the speed of GPT-5 is psychologically disorienting. “It just knows when to think,” he said. “It delivers profound responses at astonishing speed.” GPT-5 is rolling out to all tiers, including Free, Plus, Pro, and Team users. Usage limits will depend on the plan. 

OpenAI’s ChatGPT-5 debut follows its ChatGPT Agent reveal from two weeks ago. On Stocktwits, retail sentiment around the company was in the ‘extremely bullish’ zone amid ‘high’ levels of chatter.

Altman stated that while ChatGPT-3 may have been like talking to a “high school student” and GPT-4 more like a “college student”, ChatGPT-5 was like communicating with an “expert.” “With GPT-5, it’s like talking to an expert, a legitimate PhD-level expert in anything, any area you need, on demand, that can help you with whatever your goals are,” he said during the live stream.

Altman explained that GPT-5 was capable of writing an entire computer program from scratch. “We think this idea of software on demand is going to be one of the defining characteristics of the GPT-5 era,” he said, adding that GPT-5 can help people plan a party, better understand their health, and perform other tasks. “This is an incredible superpower on demand,” he said.

While explaining the updates, one of the team members said that a key area of focus for OpenAI with GPT-5 was improving reliability and being accurate with facts. He also said that it has specifically focused on reliability in health.

The team announced that one can ask GPT-5 to “think harder” or “more precisely” in the actual prompt. Users will also be able to see GPT-5's thought process while it's answering the prompt.

They explained that while OpenAI’s new model is the most powerful one yet, its real upgrade isn’t raw power but its unity. The company also focused on improvising voice and video functionality. Voice mode will be rolling out for all users with the launch of ChatGPT-5.

One of the other team members said that not only has GPT-5 been developed to mitigate hallucinations, but also to reduce deception. This means that if someone asks for information that could be potentially harmful, the model has the option to refuse an answer or offer alternatives. The feature is dubbed ‘Safe Completions.’

There were also updates around how ChatGPT-5 would impact developers. The model is much better at ‘agentic coding.’ This means it’s more autonomous, more collaborative, and better at reasoning through complex tasks.

The upgrade will also be able to create simple games if a user is using precise prompts.

Here's a quick list of all the updates that were announced:

- More accurate

- Fewer hallucinations

- Faster thinking

- Delivers clearer responses

- The best model so far for health-related guidance

- Chat color customization 

- Pre-set personalities like Cynic, Robot, Listerner, and Nerd

- Unified Voice Mode 

- Safe Completion feature

- Gmail, Google Calendar, and Contacts integration

OpenAI announced two new model variants–a lightweight GPT-5-mini and an even faster, cheaper GPT-5-nano, which is only available in the API. According to the company, free users would get access to GPT-5 and GPT-5-mini, while the Plus subscription includes the same models with “significantly higher” usage limits. 

It added that the $200 a month Pro tier offers unlimited GPT-5 access, along with GPT-5-pro and GPT-5-thinking. GPT-5 would cost developers using the API $1.25 per one million input tokens and $10 per one million output tokens. “GPT-5 mini is priced at $0.25 per 1 million input tokens and $2 per 1 million output tokens, and GPT-5 nano is priced at $0.05 per 1 million input tokens and $0.40 per 1 million output tokens,” the company announced.

Altman had earlier said that “something big was coming” on Friday, which most assumed to mean the launch of ChatGPT-5. However, OpenAI announced a “Live5tream” earlier on Thursday.

The official announcement comes after GitHub accidentally announced OpenAI’s new range of GPT-5 models on Wednesday, according to The Verge.

Read also: Jim Cramer Says Trump’s Call For Lip-Bu Tan’s Resignation Is 'Incredibly Salient' For Intel

For updates and corrections, email newsroom[at]stocktwits[dot]com.