OpenAI GPT-5.1 Eight New Personalities
- On Wednesday, OpenAI released GPT-5.1 Instant and GPT-5.1 Thinking, two updated versions of its flagship AI models now available in ChatGPT.
- The release follows complaints earlier this year that previous models were excessively cheerful and sycophantic, alongside a contrasting controversy regarding how OpenAI modified the default GPT-5 output style...
- the company now faces intense scrutiny from lawyers and regulators, potentially threatening its future operations.
“`html
On Wednesday, OpenAI released GPT-5.1 Instant and GPT-5.1 Thinking, two updated versions of its flagship AI models now available in ChatGPT. The company is framing the models with anthropomorphic language, claiming they’re warmer, more conversational, and better at following instructions.
The release follows complaints earlier this year that previous models were excessively cheerful and sycophantic, alongside a contrasting controversy regarding how OpenAI modified the default GPT-5 output style after several suicide lawsuits.
the company now faces intense scrutiny from lawyers and regulators, potentially threatening its future operations. In this habitat, simply releasing a new AI model with a few statistics is a different proposition than it was even a year ago. Here are the basics: GPT-5.1 Instant will be ChatGPT’s faster default option for most tasks, while GPT-5.1 Thinking is a simulated reasoning model designed to handle more complex problem-solving.
OpenAI claims both models outperform GPT-5 (released in August) on technical benchmarks like math and coding evaluations (including AIME 2025 and Codeforces).
Improved benchmarks may appeal to some users, but the most significant change with GPT-5.1 lies in its presentation. OpenAI states it responded to user feedback requesting AI models that…
Here’s a table summarizing the reported performance improvements:
| Benchmark | GPT-5 | GPT-5.1 Instant | GPT-5.1 Thinking |
|---|---|---|---|
| AIME 2025 (Math) | 72% | 78% | 82% |
| Codeforces (Coding) | 65% | 70% |