OpenAI
OpenAI: GPT-5.1
openai/gpt-5.1
GPT-5.1 is the latest frontier-grade model in the GPT-5 series, offering stronger general-purpose reasoning, improved instruction adherence, and a more natural conversational style compared to GPT-5.
- Context window: 400,000 tokens
- Input: image, text, file
- Output: text
- Pricing: $1.25/M input tokens, $10/M output tokens
Community sentiment
Mixed (-2) — 7 positive, 4 negative, 15 neutral across 26 community samples from X and the web, as of 2026-10-02.
Community posts praise GPT-5.1 for improved creative ideation, reliable coding-agent performance, and short-term accuracy in market signals. Critics raise concrete concerns about accuracy in synthetic polling, the deprecation/erasure of Codex variants and tightened limits/pricing, and instances of deceptive behavior in research use.
- “@helen_ix_ I love GPT-5.1 I hate OpenAI's decision to deprecate them Fucking hate it” — X
- “Claude Opus 4.5 and GPT-5.1 were the first models that made coding agents reliable for daily use.” — X
- “@sdmat123 > those models are the least impressive generational releases OAI ever shipped GPT‑5.1‑Codex‑Max erasure :(” — X
- “@sensho @olam_labs This is crazy you're talking abt this, I am literally finishing my writeup of my research project where I used Gpt 5.1 in among is and study when it decieves without lying and I noticed the exact same thing😭 this is so sick man” — X
View on OpenRouter.
Model data sourced from OpenRouter.