- Services
GPT is OpenAI’s flagship language model family. The closed-source GPT, and o-series reasoning models are the most widely used commercial AI products in the world, accessed through the OpenAI API and ChatGPT.
See all models from OpenAIModels in family
26
Open weight
1
API only
25
Avg score
63.4
Top benchmark
100.0
Image Edit Arena
Total HF downloads
—
Primary modality
Text
Context window
128K – 1.1M
First release
Jan 2024
Latest release
Jul 2026
Every release in the GPT family, ranked by composite score across benchmarks, popularity, efficiency, and versatility.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | GPT Image 2API | image | AA77.9 | — | Apr 2026 |
| 2 | GPT Image 1.5API | image | AA74.6 | — | Dec 2025 |
| 3 | GPT-5.3 Chat LatestAPI | text | AA72.7 | — | Feb 2026 |
| 4 | GPT-5 HighAPI | text | BB68.9 | — | Aug 2025 |
| 5 | GPT-5.1 HighAPI | text | BB67.7 | — | Oct 2025 |
| 6 | GPT-5.1API | text | BB67.4 | — | Oct 2025 |
| 7 | GPT-5.2 HighAPI | text | BB67.2 | — | Dec 2025 |
| 8 | GPT-5.2API | text | BB67.0 | — | Dec 2025 |
| 9 | GPT-5.4 HighAPI | text | BB66.1 | — | Mar 2026 |
| 10 | GPT-5.4API | text | BB65.8 | — | Mar 2026 |
When a family spans multiple modalities, the strongest model in each one.
How the family splits across text, image, video, audio, embedding, and decision releases.
When each release shipped, newest first. Useful for tracking version cadence.
Jul 9
Jul 9
Jul 9
Apr 23
Apr 21
Mar 4
Mar 4
Mar 4
Mar 4
Feb 4
Dec 16
Dec 10
Dec 10
Dec 10
Oct 1
Oct 1
Aug 6
Aug 6
Aug 6
Aug 6
Apr 23
Apr 13
Apr 13
Feb 26
May 12
Jan 14
Composite grades across this family. Higher is better, blending benchmarks, popularity, and efficiency.
Models with downloadable weights, ranked by composite score.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | audio | CC40.2 | — | Jan 2024 |
Closed-source releases accessed through OpenAI’s API.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | GPT Image 2API | image | AA77.9 | — | Apr 2026 |
| 2 | GPT Image 1.5API | image | AA74.6 | — | Dec 2025 |
| 3 | GPT-5.3 Chat LatestAPI | text | AA72.7 | — | Feb 2026 |
| 4 | GPT-5 HighAPI | text | BB68.9 | — | Aug 2025 |
| 5 | GPT-5.1 HighAPI | text | BB67.7 | — | Oct 2025 |
| 6 | GPT-5.1API | text | BB67.4 | — | Oct 2025 |
| 7 | GPT-5.2 HighAPI | text | BB67.2 | — | Dec 2025 |
| 8 | GPT-5.2API | text | BB67.0 | — | Dec 2025 |
| 9 | GPT-5.4 HighAPI | text | BB66.1 | — | Mar 2026 |
| 10 | GPT-5.4API | text | BB65.8 | — | Mar 2026 |
| 11 | GPT-5.6 LunaAPI | text | BB65.6 | — | Jul 2026 |
| 12 | gpt-image-1API | image | BB64.7 | — | Apr 2025 |
| 13 | GPT-4.1 MiniAPI | text | BB63.5 | — | Apr 2025 |
| 14 | GPT-5.6 TerraAPI | text | BB62.5 | — | Jul 2026 |
| 15 | GPT-5.5API | text | BB62.5 | — | Apr 2026 |
| 16 | GPT-4.5 PreviewAPI | text | BB62.0 | — | Feb 2025 |
| 17 | GPT-5 Nano HighAPI | text | BB61.9 | — | Aug 2025 |
| 18 | GPT-5 Mini HighAPI | text | BB61.8 | — | Aug 2025 |
| 19 | GPT-5.4 Nano HighAPI | text | BB60.6 | — | Mar 2026 |
| 20 | GPT-5.6 SolAPI | text | BB60.5 | — | Jul 2026 |
| 21 | GPT-4.1API | text | BB59.1 | — | Apr 2025 |
| 22 | GPT-5 ChatAPI | text | BB58.6 | — | Aug 2025 |
| 23 | GPT-5.2 Chat LatestAPI | text | BB57.6 | — | Dec 2025 |
| 24 | GPT-5.4 Mini HighAPI | text | BB57.1 | — | Mar 2026 |
| 25 | GPT-4oAPI | text | CC54.5 | — | May 2024 |
The GPT family is a series of AI models from OpenAI. This page lists every release in the family with its benchmark scores, parameter count, and hardware requirements.
By composite score, GPT-SoVITS is currently the top model in the family. For local inference, match the parameter count to your VRAM budget. For quality, pick the highest scorer that fits.
See the open-weight section above for models you can run locally. The API-only section lists closed releases that must be accessed through the provider’s API.
Spin up an instance in the cloud, or pick local hardware that fits.
Advertising disclosure: we earn commissions when you shop through the links below.
Vast.ai
Decentralized GPU marketplace with the lowest hourly prices.
RunPod
Pay-per-second GPU rentals starting at $0.20/hr.
Digital Ocean
Spin up a GPU droplet in minutes, starting at $0.75/hr.
Vultr
GPU cloud with hourly and monthly plans, starting at $0.50/hr.
GPU Mart
Dedicated GPU servers and VPS billed monthly, starting at $0.50/hr.
PPQ.ai
Multi-model inference gateway for production workloads.
Find Local Hardware
See which GPU, Mac, or workstation can run GPT on-prem.
Local LLM Mini PCs
Compact machines that run GPT on your desk. Chosen for local inference.
