OpenAI
United States · Global · 4 model families / 3 agent tools
Headquartered in United States. This index covers 4 model families / 24 versions and 3 agent tools in Image / Language & multimodal / Speech & audio / Video, 6 of them active and 4 with open weights. Lineups are listed by category below, with a full lineage ordered by official release date.
Families4Versions24Agents3CategoriesImage / Language & multimodal / Speech & audio / VideoActive / open6 / 4Bench records119Updated2026-11-08
01 / Models
Language & multimodal
1 entries| Model | Status | Openness | versions | Last updated | |
|---|---|---|---|---|---|
| GPT | Active | Closed source | 9 |
02 / Models
Image
1 entries| Model | Status | Openness | versions | Last updated | |
|---|---|---|---|---|---|
| GPT Image (DALL·E lineage) | Active | Closed source | 8 |
03 / Models
Video
1 entries| Model | Status | Openness | versions | Last updated | |
|---|---|---|---|---|---|
| Sora | Discontinued | Closed source | 3 |
04 / Models
Speech & audio
1 entries| Model | Status | Openness | versions | Last updated | |
|---|---|---|---|---|---|
| Whisper | Active | Open source | 4 |
05
Agents
06
Version lineage
- GPT Image 2.6 · GPT Image (DALL·E lineage) Image TBD
- GPT-6.1 Terra · GPT Language & multimodal TBD
- GPT-6.1 Sol · GPT Language & multimodal TBD
- GPT-6 Sol · GPT Language & multimodal Superseded TBD
- GPT Image 2.5 · GPT Image (DALL·E lineage) Image Superseded TBD
- GPT-6 Astra · GPT Language & multimodal TBD
- GPT-5.6 Sol · GPT Language & multimodal Superseded TBD
- GPT Image 2 · GPT Image (DALL·E lineage) Image Superseded TBD
- GPT Image 1.5 · GPT Image (DALL·E lineage) Image Superseded TBD
- Sora 2 · Sora Video Discontinued TBD
- GPT-5 · GPT Language & multimodal Superseded TBD
- GPT-4.1 · GPT Language & multimodal Superseded TBD
- GPT Image 1 · GPT Image (DALL·E lineage) Image Superseded TBD
- Sora Turbo (Sora 1) · Sora Video Discontinued TBD
- large-v3-turbo · Whisper Speech & audio Speed-optimised cut of large-v3: decoder pruned to 4 layers, ~809M parameters, several times faster at near-equal accuracy.
- GPT-4o · GPT Language & multimodal Superseded TBD
- Sora（研究预览） · Sora Video Discontinued TBD
- large-v3 · Whisper Speech & audio Released at OpenAI DevDay 2023: 128 Mel bins, added Cantonese token, trained on ~5M hours including pseudo-labelled audio.
- DALL·E 3 · GPT Image (DALL·E lineage) Image Discontinued TBD
- GPT-4 · GPT Language & multimodal Discontinued TBD
- large-v2 · Whisper Speech & audio Superseded Improved training and regularisation over the original large model.
- Whisper 首发（tiny→large 五档） · Whisper Speech & audio Superseded Open-sourced September 2022, trained on ~680k hours of weakly supervised audio, code and weights under MIT.
- DALL·E 2 · GPT Image (DALL·E lineage) Image Discontinued TBD
- DALL·E · GPT Image (DALL·E lineage) Image Superseded TBD