← AI news, last 7 days

GPT-4o

A model by OpenAI · released 2024-05-13 · 75 checked facts from 16 sources

Timeline

Prices

Per price, each value with the dates it was seen. A value seen on one date only is one copy of the page. 1 price changed over the record.
Price forValues over time
per 1M cached input tokens (audio, realtime)$2.50 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M cached input tokens (fine-tuning)$1.875 2025-05-03 · OpenAI's page as archived 2025-05-03
per 1M cached input tokens (standard)$1.25 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M cached input tokens (text, realtime)$2.50 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M input tokens (audio, realtime)$40 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M input tokens (batch)$1.25 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M input tokens (fine-tuning)$3.75 2025-05-03 · OpenAI's page as archived 2025-05-03
per 1M input tokens (standard)$5 2024-08-01 · OpenAI's page as archived 2024-08-01 → $2.50 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M input tokens (text, realtime)$5 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M output tokens (audio, realtime)$80 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M output tokens (batch)$5 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M output tokens (fine-tuning)$15 2025-05-03 · OpenAI's page as archived 2025-05-03
per 1M output tokens (standard)$10 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M output tokens (text, realtime)$20 2025-05-03 to 2025-08-02 · OpenAI's page as archived 2025-05-03, OpenAI's page as archived 2025-08-02
per 1M tokens (fine-tuning training)$25 2025-05-03 · OpenAI's page as archived 2025-05-03
per 1M tokens (vision, fixed)$5 2024-08-01 · OpenAI's page as archived 2024-08-01
image (low resolution)$0.001275 2024-08-01 · OpenAI's page as archived 2024-08-01

Dated versions

Pinned snapshots of the model, priced on their own.

gpt-4o-2024-05-13

per 1M input tokens (standard)$5 2024-08-01 to 2025-02-01 · OpenAI's page as archived 2024-08-01, OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M output tokens (batch)$7.50 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M output tokens (standard)$15 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01

gpt-4o-2024-08-06

per 1M cached input tokens (fine-tuning standard)$1.875 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M input tokens (batch)$1.875 2025-02-01 · OpenAI's page as archived 2025-02-01
per 1M input tokens (fine-tuning batch)$1.875 2024-11-12 · OpenAI's page as archived 2024-11-12
per 1M input tokens (fine-tuning standard)$3.75 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M output tokens (batch)$7.50 2025-02-01 · OpenAI's page as archived 2025-02-01
per 1M output tokens (fine-tuning batch)$7.50 2024-11-12 · OpenAI's page as archived 2024-11-12
per 1M output tokens (fine-tuning standard)$15 2024-11-12 to 2025-02-01 · OpenAI's page as archived 2024-11-12, OpenAI's page as archived 2025-02-01
per 1M tokens (fine-tuning training)$25 2024-11-12 · OpenAI's page as archived 2024-11-12
per 1M tokens (training)$25 2025-02-01 · OpenAI's page as archived 2025-02-01

Scores

Self-reported: stated on the lab's own channel. A score reported twice with two values is shown twice, with its date: a lab re-runs a benchmark under a new setup; neither replaces the other.
BenchmarkScoreDateWho
AIME '2413.1%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
AIME ‘24 (math)9.3%2025-02-28self-reported · OpenAI's page as archived 2025-02-28
AIME’249.3%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
ARC-AGI50%2026-10-01reported by stoppels.ch · stoppels.ch
Aider’s polyglot18.2%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
CharXiv-D85.3%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
CharXiv-Reasoning52.7%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
EgoSchema72.2%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
GPQA (science)53.6%2025-02-19self-reported · xAI's page as archived 2025-02-19, OpenAI's page as archived 2025-02-28
GPQA Diamond46.0%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
Graphwalks41.7%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
IFEval81.0%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
International Mathematics Olympiad (IMO)132024-09-12self-reported · OpenAI's page as archived 2024-09-12
LOFT (128k)78.0%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
LiveCodeBench32.3%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
MMLU87.2%2024-05-13self-reported · OpenAI's page as archived 2024-05-13
MMLU85.7%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
MMLU-pro72.6%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
MMMLU (multilingual)81.5%2025-02-28self-reported · OpenAI's page as archived 2025-02-28
MMMU69.1%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
MMMU68.7%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
MMMU (multimodal)69.1%2025-02-28self-reported · OpenAI's page as archived 2025-02-28
MathVista61.4%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
MultiChallenge27.8%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
Multilingual MMLU81.4%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
SWE-Bench Verified (coding)30.7%2025-02-28self-reported · OpenAI's page as archived 2025-02-28
SWE-Lancer Diamond (coding)23.3%2025-02-28self-reported · OpenAI's page as archived 2025-02-28
SWE-bench Verified33.2%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
Script concordance testing (SCT)63.9%2026-10-26reported by PubMed: AI in top medical journals · PubMed: AI in top medical journals
SimpleQA38.2%2025-02-19reported by xAI · xAI's page as archived 2025-02-19
Video-MME65.3%2025-04-14self-reported · OpenAI's page as archived 2025-04-14
jailbreaking test222024-09-12self-reported · OpenAI's page as archived 2024-09-12

What it does

In the news

Built 2026-11-21 06:50 UTC from the news knowledge graph by code. Every line is a fact whose quote was checked against the source's own words (hover a source); it says who states it, not that it is true. Archived pages are the lab's pages as the Wayback Machine holds them.