OpenAI

GPT 5 2025-08-07

Type
language model
Context
272K tokens
Max output
128K
API string
azure/eu/gpt-5-2025-08-07

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
Microsoft Foundry azure/gpt-5-2025-08-07 $1.25 $10 31 Jul 2026
OpenAI $1.25 $10 31 Jul 2026
Microsoft Foundry azure/eu/gpt-5-2025-08-07 $1.375 $11 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

A single provider appears several times if it sells this model under different identifiers and at different prices — for example, at different weight precision. The identifier is shown next to the name. Identical offers that arrived from two sources under different spellings are merged into one row.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
MATH, difficulty level five 98.13 % effort: high · with a tuned harness Epoch evaluations
Fiction.LiveBench — holding a long context 97.2 % effort: medium Fiction.live leaderboard
Mock AIME 2024–2025 — olympiad problems 91.38 % effort: minimal · with a tuned harness Epoch evaluations
Aider Polyglot — code edits in six languages 88 % effort: low · with a tuned harness Aider LLM Leaderboards
Creative writing (Lech Mazur’s evaluation) 86 % effort: medium lechmazur/writing Github repository
GPQA Diamond — graduate-level questions 81.57 % effort: minimal · with a tuned harness Epoch evaluations
GeoBench — locating a place from a photograph 81 % effort: medium GeoBench leaderboard
SWE-bench Verified — fixing bugs in repositories 73.55 % effort: high · with a tuned harness Epoch evaluations
ARC-AGI — generalising to unseen patterns 65.7 % effort: high · with a tuned harness ARC Prize Leaderboard
WeirdML — unusual machine learning tasks 60.7 % effort: high WeirdML Leaderboard
FrontierMath, levels 1–3 55.44 % effort: high Epoch evaluations
DeepResearch Bench — deep research 55.13 % DeepResearchBench Leaderboard
SimpleQA Verified — factual accuracy 50.6 % effort: high Epoch evaluations
Terminal-Bench — working in the command line 49.6 % · with a tuned harness Terminal-Bench v2 Leaderboard
SimpleBench — trick questions 48.04 % effort: high SimpleBench Leaderboard
GDPval — tasks from real occupations 34.8 % effort: medium
Chess puzzles 33.71 % effort: minimal Epoch evaluations
BALROG — game environments 32.8 % effort: minimal Balrog Leaderboard
FrontierMath, level 4 — research-grade problems 21.95 % effort: high Epoch evaluations
Humanity’s Last Exam — expert-level questions 21.56 % effort: high
APEX-Agents 18.3 % effort: high
CritPt — physics problems 12.6 % effort: high
ARC-AGI-2 9.86 % effort: high
GSO-Bench — code optimisation 6.9 % effort: high GSO Leaderboard
Remote Labor Index — jobs from a freelance marketplace 1.67 %