whichllm — Browse and compare AI model specs and pricing

Poe

GPT-5.4

models.dev synced record

Model ID openai/gpt-5.4
Provider Poe
Family -
Status -
Knowledge Cutoff -
Release Date 2026-02-26
Input Modalities text, image, pdf
Output Modalities image
Context Window 1050000
Input Limit 922000
Output Limit 128000
Tool Calling Yes
Reasoning Yes
Structured Output -
Temperature Control No
Open Weights No
Input Cost / 1M tokens $2.20
Output Cost / 1M tokens $14.00
Reasoning Cost / 1M tokens -
Cache Read Cost / 1M tokens $0.22
Cache Write Cost / 1M tokens -