AA Intelligence Index v4.3: DeepSeek V4 Pro Edges Qwen 27B as Flash Next Beats Its Own MoE Flagship
pbaylies · x · 2026-09-08
Artificial Analysis shipped Intelligence Index v4.3, upgrading Terminal-Bench to 4.0 and replacing τ³-Banking with AutomationBench-AA, a private-set agentic workflow automation benchmark, as a bridge toward Index v5.
Key leaderboard shifts:
- DeepSeek V4 Pro now beats Qwen 3.8 27B (xhigh) by 2 points (36 vs 34, previously tied at 42)
- Qwen 3.8 Flash Next's lead over its own 27B (xhigh) doubled to 8 points
- Flash Next (42) flipped past Gemini 3.8 Flash (41), which previously led by 1
- Flash Next beats Qwen's own massive 2.4T A95B MoE (42 vs 40)
- Reasoning-compression on 27B Dense: the Low→xHigh spread shrank from 8 points (34→42) to 5 (29→34)
The takeaway: small, fast Flash Next is outperforming far larger models, and returns from higher reasoning effort are compressing.
Related event: Artificial Analysis Rolls Out Intelligence Index v4.3 with New Benchmarks(8 posts)→
More from Models
- GPT-6.1 Terra rolls out to all Plus, Pro, Enterprise and Edu users — OpenAI · 2026-10-27
- OpenAI launches GPT-6.1 Terra with 50% lower API prices than GPT-5.6 Terra — OpenAI · 2026-10-27
- GPT-6.1 Terra API pricing revealed: $1 input / $5 output, $0.05 cached — airesearch12 · 2026-10-27
- In cyber, intention matters little: Chintala on model info leaks, and what users should do — brhydon · 2026-10-27
- Why don't more LLMs ship 10M-token context? A researcher asks about attention tradeoffs — agarwl_ · 2026-10-27
- Ortet's first health eval: 34 open-weight models, 50 tasks, generic rankings fall short — thegautamkamath · 2026-10-27