Pass IndexThe State of AISign in

GPT-5 Pro

GPT-5 pro uses more compute to think harder and provide consistently better answers.

text + image + file → text · made by OpenAI

$15$120per Mtok in / out
OpenAI · 6 sellers

Sold by 9 ways

SellerLaneRate
Microsoft Azure AI cloudbatch$7.5per Mtok in$60per Mtok outprices.azure.com · read 2026-08-24
Microsoft Azure AI cloudstandard$15per Mtok in$120per Mtok outmodels.dev · read 2026-09-16
OpenAI apistandard$15per Mtok in$120per Mtok outdevelopers.openai.com · read 2026-08-24
Nous Research 2 ways$15per Mtok in$120per Mtok out
Nous Research aggregatorbatch$7.5per Mtok in$60per Mtok outinference-api.nousresearch.com · read 2026-09-12
Nous Research aggregatorstandard$15per Mtok in$120per Mtok outinference-api.nousresearch.com · read 2026-09-12
OpenRouter 2 ways$15per Mtok in$120per Mtok out
OpenRouter aggregatorbatch$7.5per Mtok in$60per Mtok outopenrouter.ai · read 2026-08-24
OpenRouter aggregatorstandard$15per Mtok in$120per Mtok outopenrouter.ai · read 2026-08-24
Requesty aggregatoropenai-responses$15per Mtok in$120per Mtok out$15per Mtok cachedrouter.requesty.ai · read 2026-08-28
Vercel AI Gateway aggregatorstandard$15per Mtok in$120per Mtok outai-gateway.vercel.sh · read 2026-08-25

Measured 15 standings

PlaceBoardMetricScore
7thof 46Enigma eval — Epoch AIAccuracy18.75
9thof 51Humanity's Last Exam (Epoch AI replication)Accuracy31.64
18thof 18Algotune — Epoch AIScore (medium)1.31
19thof 72Frontiermath tier 4 — Epoch AImean_score (high)14.6
25thof 102Simplebench — Epoch AIScore (AVG@5) (high)0.616
35thof 106Frontiermath tiers 1 3 v2 — Epoch AImean_score (high)55.789
42ndof 62Frontiermath tier 4 v2 — Epoch AImean_score (high)19.512
48thof 170Weirdml — Epoch AIAccuracy (high)60.39
94thof 205ARC-AGI-1 (Public Eval)Score (%)77
100thof 216ARC-AGI-2 (Public Eval)Score (%)13.33
106thof 232ARC-AGI-2 (Semi-Private)Score (%)18.33
107thof 230ARC-AGI-1 (Semi-Private)Score (%)70.17
110thof 227Arc agi 2 — Epoch AIScore18.33
116thof 244Arc agi — Epoch AIScore (high)70.2
202ndof 656Epoch capabilities index — Epoch AIECI Score (high)150.3

About

GPT-5 Pro — a text model from OpenAI, sold by 6 companies from $15 in and $120 out per million tokens, placed 7th of 46 on Enigma eval — Epoch AI.

It takes text, images and files and returns text, with a context window of 400,000 tokens. It was published in 2025, trained on material up to September 2024. Its sellers say it can reason step by step and call a tool. The catalogue files it under extract and reasoning. Six companies sell it. The cheapest is $15 in and $120 out per million tokens at Microsoft Azure AI. Beside the standard rate there are batch and separately routed lanes. It has been measured on 15 boards, and stands best at 7th of 46 on Enigma eval — Epoch AI.

Every current figure

Maker
OpenAI
Register
model
Takes
text + image + file
Returns
text
Context
400,000 tokens
Longest answer
272,000 tokens
Published
2025
Knowledge to
September 2024
Licence
not read
Sellers
6
Price
$15 in and $120 out per million tokens — Microsoft Azure AI
Boards
15
Best place
7th of 46 — Enigma eval — Epoch AI

Known as 7 names

GPT-5 Pro (unknown thinking)gpt 5 proopenai/gpt-5-proopenai/gpt-5-pro:batchgpt-5-proopenai-responses/gpt-5-proOpenAI: GPT-5 Pro (batch)