AA Intelligence Index v4.3.2
Intelligence index (legacy) · 14 models · source Artificial Analysis · updated 2026-09
What this benchmark measures
The pre-v5 version, which covers more models including discontinued ones. Historical reference only; never mixed into v5 columns.
Why it gets its own table
It is kept because many open models still only have v4.3.2 results. For ranking the open camp against itself, this table is more complete.
Full ranking
| # | Model | Score | Vs. best | Input $ | Output $ | Type |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5Anthropic | 57.6 | Best overall | $4.00 | $20.00 | |
| 2 | Claude Sonnet 5.5Anthropic | 56.0 | 97% | $2.00 | $10.00 | |
| 3 | Kimi K3.1 Pro Open weightsMoonshot AI | 55.1 | 96% | $2.00 | $10.00 | Open weights |
| 4 | Claude Fable 5.1Anthropic | 53.0 | 92% | $10.00 | $50.00 | |
| 5 | Grok 4.7xAI | 46.4 | 81% | $2.00 | $6.00 | |
| 6 | MiMo-V2.6 Pro Open weightsXiaomi | 46.3 | 80% | $0.35 | $1.40 | Open weights |
| 7 | DeepSeek V4.1 Flash Open weightsDeepSeek | 45.1 | 78% | $0.30 | $1.20 | Open weights |
| 8 | GLM-5.3 Open weightsZ.ai | 44.8 | 78% | $0.60 | $2.20 | Open weights |
| 9 | Kimi K3 Open weightsMoonshot AI | 43.6 | 76% | $2.00 | $15.00 | Open weights |
| 10 | Claude Haiku 5.5Anthropic | 43.0 | 75% | $0.10 | $0.50 | |
| 11 | GLM-5.3 Flash Open weightsZ.ai | 41.8 | 73% | $0.20 | $0.80 | Open weights |
| 12 | Qwen3.8 27B Open weightsAlibaba Qwen | 33.7 | 59% | $0.10 | $0.40 | Open weights |
| 13 | MiniMax M3 Open weightsMiniMax | 29.2 | 51% | $0.25 | $1.00 | Open weights |
| 14 | Gemma 4 31B Open weightsGoogle | 19.0 | 33% | $0.10 | $0.40 | Open weights |