Model roster · 45 tracked systems
Every major contender, on one field.
Anthropic, OpenAI, Google, Moonshot, Alibaba, Meta, xAI, Zhipu, DeepSeek, MiniMax, Mistral, Microsoft, 01.AI and Baichuan — every comparable ECI record, plotted and ranked.
- Models
- 45
- Labs
- 14
- Open weights
- 24/ 45
- Newest release
- Mistral Large 42026-10-06
Showing 45 of 45 tracked models
| Rank | Model | Ecosystem | Access | Released | ECI |
|---|---|---|---|---|---|
| 1 | Claude Opus 5.5Anthropic | Western | API | 2026-09-22 | 166.8 |
| 2 | GPT-6 AstraOpenAI | Western | API | 2026-09-03 | 166.4 |
| 3 | GPT-6.1 SolOpenAI | Western | API | 2026-09-29 | 166.3 |
| 4 | Claude Fable 5.1Anthropic | Western | API | 2026-09-01 | 164.7 |
| 5 | Gemini 4 ArgonGoogle DeepMind | Western | Limited access | 2026-09-30 | 164.6 |
| 6 | Claude Sonnet 5.5Anthropic | Western | API | 2026-09-28 | 164.5 |
| 8 | GPT-6 SolOpenAI | Western | API | 2026-09-22 | 162.6 |
| 17 | Kimi K3Moonshot AI | Chinese | Open weightsnon-commercial | 2026-07-16 | 157.4 |
| 18 | Gemini 3.7 FlashGoogle DeepMind | Western | API | 2026-08-13 | 157.2 |
| 19 | Gemini 3.8 FlashGoogle DeepMind | Western | API | 2026-09-02 | 157.2 |
| 20 | Qwen 3.8 MaxAlibaba / Qwen | Chinese | API | 2026-08-02 | 156.9 |
| 22 | Muse Spark 1.3Meta AI | Western | API | 2026-09-02 | 156.8 |
| 28 | Grok 4.6xAI | Western | API | 2026-08-12 | 156.3 |
| 30 | GLM-5.3Zhipu AI / Z.ai | Chinese | API | 2026-08-14 | 155.9 |
| 31 | DeepSeek V4.1 FlashDeepSeek | Chinese | Open weightsunrestricted | 2026-09-09 | 155.6 |
| 32 | DeepSeek V4 Pro 0813DeepSeek | Chinese | Open weightsunrestricted | 2026-08-13 | 155.3 |
| 34 | Qwen3.8 Max (0902)Alibaba / Qwen | Chinese | API | 2026-09-01 | 154.9 |
| 35 | Muse Spark 1.2Meta AI | Western | API | 2026-08-05 | 154.9 |
| 37 | DeepSeek V4 Flash 0731DeepSeek | Chinese | Open weightsunrestricted | 2026-07-31 | 154.5 |
| 39 | Grok 4.7xAI | Western | API | 2026-09-21 | 154.4 |
| 41 | Muse Spark 1.1Meta AI | Western | API | 2026-07-09 | 154.0 |
| 42 | Grok 4.5xAI | Western | API | 2026-07-08 | 153.9 |
| 43 | Qwen3.7-MaxAlibaba / Qwen | Chinese | API | 2026-05-19 | 153.8 |
| 45 | GLM-5.3-FlashZhipu AI / Z.ai | Chinese | Open weightsunrestricted | 2026-08-20 | 153.2 |
| 50 | GLM-5.2Zhipu AI / Z.ai | Chinese | Open weightsunrestricted | 2026-06-16 | 151.9 |
| 52 | Qwen 3.8 27BAlibaba / Qwen | Chinese | Open weightsunrestricted | 2026-08-14 | 151.6 |
| 53 | Kimi K2.6Moonshot AI | Chinese | Open weightsunrestricted | 2026-04-20 | 150.9 |
| 59 | Kimi K2.7 CodeMoonshot AI | Chinese | Open weightsunrestricted | 2026-06-12 | 149.9 |
| 70 | MiniMax-M3MiniMax | Chinese | Open weightsrestricted use | 2026-06-01 | 147.1 |
| 71 | Mistral Large 4Mistral AI | Western | Unknown | 2026-10-06 | 146.9 |
| 74 | MiniMax-M2.5MiniMax | Chinese | Open weightsunrestricted | 2026-02-12 | 146.7 |
| 83 | MiniMax-M2.7MiniMax | Chinese | Open weightsnon-commercial | 2026-03-18 | 145.9 |
| 102 | Gemma 4 31B ITGoogle DeepMind | Western | Open weightsrestricted use | 2026-04-02 | 142.9 |
| 113 | Mistral Medium 3.5Mistral AI | Western | Open weightsrestricted use | 2026-04-28 | 141.5 |
| 120 | gpt-oss-120bOpenAI | Western | Open weightsunrestricted | 2025-08-05 | 140.0 |
| 149 | Mistral Medium 3Mistral AI | Western | API | 2025-05-07 | 134.1 |
| 155 | Llama 4 MaverickMeta AI | Western | Open weightsrestricted use | 2025-04-06 | 132.2 |
| 160 | Phi-4Microsoft | Western | Open weightsunrestricted | 2024-12-12 | 130.4 |
| 196 | phi-3-small 7.4BMicrosoft | Western | Open weightsunrestricted | 2024-04-23 | 121.9 |
| 197 | phi-3-medium 14BMicrosoft | Western | Open weightsunrestricted | 2024-04-23 | 121.3 |
| 211 | Yi-34B01.AI | Chinese | Open weightsrestricted use | 2023-11-02 | 117.4 |
| 245 | Yi 6B01.AI | Chinese | Open weightsrestricted use | 2023-11-22 | 104.9 |
| 247 | Baichuan2-13BBaichuan | Chinese | Open weightsrestricted use | 2023-09-06 | 103.3 |
| 257 | Baichuan 2-7BBaichuan | Chinese | Open weightsrestricted use | 2023-09-20 | 96.5 |
| 269 | Baichuan1-7BBaichuan | Chinese | Open weightsnon-commercial | 2023-06-01 | 90.6 |
No tracked model matches that. Try a lab name, like “DeepSeek”.
“Open weights” follows the source’s access label. It does not automatically mean unrestricted, commercially usable, or fully open-source. Rank is the model’s position among all scored source rows.
Coverage gaps
Popular labs still missing a comparable score.
- ByteDanceChinese
No valid scored ECI row in the current public dataset. The lab stays in scope and appears when comparable data does.
- XiaomiChinese
No valid scored ECI row in the current public dataset. The lab stays in scope and appears when comparable data does.
- BaiduChinese
No valid scored ECI row in the current public dataset. The lab stays in scope and appears when comparable data does.
- TencentChinese
No valid scored ECI row in the current public dataset. The lab stays in scope and appears when comparable data does.