Type ranking
Best New AI Models in October 2026
As of October 2026, GPT-6 Astra leads the Noometry Index with a score of 70.8, ahead of Claude Fable 5.1 at 69.0.
Last verified
Ranked models released in the last 90 days.
- GPT-6 Astra 70.8
- Claude Fable 5.1 69.0
- Claude Opus 5.5 68.6
- Claude Opus 5 67.8
- GPT-6.1 Sol 65.6
- Claude Sonnet 5.5 61.9
- Gemini 3.8 Flash 61.8
- GPT-6 Sol 61.8
- Gemini 3.7 Flash 59.8
- Kimi K3 59.5
Full ranking
36 models
| 1 | OpenAI GPT-6 Astra | OpenAI | Proprietary | 1.05M | 70.8 | $10 | $50 | 3.54 | — |
| 2 | Anthropic Claude Fable 5.1 | Anthropic | Proprietary | 1M | 69.0 | $10 | $50 | 3.45 | — |
| 3 | Anthropic Claude Opus 5.5 | Anthropic | Proprietary | 1M | 68.6 | $4 | $20 | 8.57 | — |
| 4 | Anthropic Claude Opus 5 | Anthropic | Proprietary | 1M | 67.8 | $5 | $25 | 6.78 | — |
| 5 | OpenAI GPT-6.1 Sol | OpenAI | Proprietary | 1.05M | 65.6 | $2 | $10 | 16.4 | — |
| 6 | Anthropic Claude Sonnet 5.5 | Anthropic | Proprietary | 1M | 61.9 | $2 | $10 | 15.5 | — |
| 7 |  Google Gemini 3.8 Flash | Proprietary | 1.05M | 61.8 | $0.75 | $3.75 | 41.2 | — | |
| 8 | OpenAI GPT-6 Sol | OpenAI | Proprietary | 1.05M | 61.8 | $2 | $10 | 15.4 | — |
| 9 | Google Gemini 3.7 Flash | Proprietary | 1.05M | 59.8 | $0.75 | $3.75 | 39.9 | — | |
| 10 | Moonshot AI Kimi K3 | Moonshot AI | Open | 1.05M | 59.5 | $3 | $15 | 9.92 | — |
| 11 | xAI Grok 4.6 | xAI | Proprietary | 500K | 56.9 | $2 | $6 | 19 | — |
| 12 |  Alibaba (Qwen) Qwen3.8 Max | Alibaba (Qwen) | Proprietary | 1M | 56.8 | $2 | $6 | 18.9 | — |
| 13 | Google Gemini 4 Argon | Proprietary | — | 56.5 | — | — | — | — | |
| 14 | Z.ai (Zhipu) GLM-5.3 | Z.ai (Zhipu) | Open | 1M | 54.8 | $1.40 | $4.40 | 25.5 | — |
| 15 |  Meta Muse Spark 1.3 | Meta | Proprietary | 1.05M | 54.8 | $1.25 | $4.25 | 27.4 | — |
| 16 | Google Gemini 3.6 Flash | Proprietary | 1.05M | 54.1 | $0.75 | $3.75 | 36.1 | — | |
| 17 | OpenAI GPT-6 Luna | OpenAI | Proprietary | 1.05M | 53.3 | $0.10 | $0.50 | 266 | — |
| 18 | xAI Grok 4.7 | xAI | Proprietary | 500K | 53.1 | $2 | $6 | 17.7 | — |
| 19 |  DeepSeek DeepSeek V4.1 Flash | DeepSeek | Open | 1M | 52.8 | $0.15 | $0.60 | 201 | — |
| 20 | Z.ai (Zhipu) GLM-5.3-Flash | Z.ai (Zhipu) | Open | 1M | 51.8 | $0.15 | $0.50 | 218 | — |
| 21 | Meta Muse Spark 1.2 | Meta | Proprietary | 1.05M | 50.3 | $1.25 | $4.25 | 25.2 | — |
| 22 | Xiaomi MiMo-V2.6-Pro | Xiaomi | Open | 1.05M | 50.3 | $0.43 | $0.87 | 92.5 | — |
| 23 | Anthropic Claude Haiku 5.5 | Anthropic | Proprietary | 1M | 49.5 | $0.10 | $0.50 | 247 | — |
| 24 | Xiaomi MiMo-V2.6-Flash | Xiaomi | Open | 1.05M | 48.5 | $0.14 | $0.28 | 277 | — |
| 25 |  StepFun Step 5 Preview | StepFun | Proprietary | 1.02M | 47.9 | $1 | $2.70 | 33.6 | — |
| 26 | Thinking Machines Lab Inkling-Small | Thinking Machines Lab | Open | 524K | 46.5 | $0.45 | $1.20 | 72.9 | — |
| 27 | Alibaba (Qwen) Qwen3.8 27B | Alibaba (Qwen) | Open | 262K | 46.0 | $0.99 | $1.49 | 41.2 | — |
| 28 |  Tencent Hy4 preview | Tencent | Open | 1.05M | 45.3 | $0.75 | $2.25 | 40.2 | — |
| 29 | Thinking Machines Lab Inkling | Thinking Machines Lab | Open | 66K | 44.1 | $1.87 | $4.68 | 17.1 | — |
| 30 |  Mistral AI Mistral Large 4 | Mistral AI | Proprietary | 1.05M | 43.1 | $0.68 | $2.09 | 41.7 | — |
| 31 |  Upstage Solar Pro4 | Upstage | Proprietary | 524K | 42.1 | $0.30 | $1.20 | 80.2 | — |
| 32 | Meta Muse Glimmer | Meta | Open | — | 41.7 | — | — | — | — |
| 33 | Google Gemini 3.5 Flash Lite | Proprietary | 1.05M | 41.5 | $0.30 | $2.50 | 48.8 | — | |
| 34 |  NVIDIA Nemotron 3.5 Lightning | NVIDIA | Open | 262K | 40.0 | $0.05 | $0.20 | 457 | — |
| 35 | Alibaba (Qwen) Qwen3.7 Flash | Alibaba (Qwen) | Proprietary | 1M | 39.9 | $0.03 | $0.13 | 726 | — |
| 36 | Inception Mercury 2.5 | Inception | Proprietary | 260K | 33.5 | $0.04 | $0.15 | 496 | — |
Sponsored placements are available on pages like this one. Advertise on Noometry
Compare the leaders
Frequently asked questions
What are the best AI models released in the last 90 days?
As of October 2026, GPT-6 Astra leads the Noometry Index with a score of 70.8, ahead of Claude Fable 5.1 at 69.0.
What are the top 5 in this ranking?
GPT-6 Astra, Claude Fable 5.1, Claude Opus 5.5, Claude Opus 5, GPT-6.1 Sol, in that order, as of October 2026.
How is this list ranked?
By the Noometry Index, which is fitted from every public benchmark result and adjusted for how hard each benchmark is. Models need several independent results to be ranked.