ClosedOpenAI/10 trustVerified

GPT-4o

Name: GPT-4o
Rating: 7

OpenAI's omni multimodal flagship: natively text, vision, and audio in one model.

Visit official page →

About

GPT-4o ("o" for omni), released May 13, 2024, was OpenAI's first natively multimodal flagship: a single model accepting and generating text, images, and audio — including real-time voice-to-voice conversation via Advanced Voice Mode (Sept 2024) and the Realtime API (Oct 2024). At release it set records on audio speech recognition/translation and vision benchmarks, scored 88.7 on MMLU, and covered 50+ languages. 128K-token context, proprietary/closed weights, available via the OpenAI API, Azure OpenAI, and ChatGPT (free tier with limits; Plus gets 5x higher limits). Fine-tuning for enterprise customers arrived Aug 2024.

Specs

Context window

128K tokens

Pricing

Usage-based API pricing (input/output per 1M tokens); at release $2.50/$10 per 1M — verify the live OpenAI pricing page, as rates and model snapshots have changed. ChatGPT: free tier with limits, Plus ($20/mo) for higher limits.

Free quota

ChatGPT free tier includes limited GPT-4o access; API has no free tier.

API access

OpenAI API, Azure OpenAI Service, ChatGPT. No weights, no self-host.

Weights

Closed — API only

Released

May 13, 2024

Benchmarks

At release (OpenAI, May 2024): 88.7 MMLU (vs 86.5 GPT-4), SOTA audio speech recognition and translation, SOTA vision benchmarks. Independent comparisons since (e.g. vs Claude 3.5 Sonnet): roughly tied on knowledge (88.7/88.7 MMLU), Sonnet ahead on GPQA/GSM8K, GPT-4o ahead on MATH — treat 2024 figures as historical; the model has been superseded by GPT-5.x for frontier work.

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

chatvisionvoicemultimodal

Trust score — 7/10

/10 trustVerified

Scored 7/10 (verified tier, metrics-v1). Evidence: provider-tier estimate — thinner evidence (no public adoption metric); released 883 days ago. Manual review: defining 2024-25 multimodal model (native text/vision/audio, 128K, MMLU 88.7 at release); still widely deployed; pricing shifted over time — verify live. License: Proprietary — closed weights.

How the trust score is built

Full methodology →

Adoption · stars/downloads, log-scaled vs peers: 9.0/10
Maintenance · recency, releases, issue load: 4.0/10
Openness · license + docs: 4.4/10
Manual review · sanity read + red-flag check: 9.0/10

· provider-tier estimate — thinner evidence (no public adoption metric)
· released 883 days ago
· manual review: defining 2024-25 multimodal model (native text/vision/audio, 128K, MMLU 88.7 at release); still widely deployed; pricing shifted over time — verify live