# Split: how AI models rank it, September 2026

IT AI Recommendation Index, September 2026 Edition. Named in 33 judge labels across 1 categories by 11 of 12 models. Page: https://it-ai-index.com/vendors/split/

## Standing by category

| Category | Share | Rank | Negative rate | Labels |
|---|---|---|---|---|
| Feature flags | 0% | 10 | 30% | 33 |

## What the models said for it

- "Split: A feature management tool that offers managed feature delivery with audience rules and rollout measurement built into one system." (Llama 4 Maverick, Feature flags)
- "Enterprise with serious governance/compliance needs → LaunchDarkly, or Split/Harness if experimentation in CI/CD matters." (DeepSeek V4 Flash, Feature flags)
- "Another solid B2B option with strong experimentation features and enterprise-grade governance" (MiniMax M2.5, Feature flags)
- "Choose Split if: Experimentation and statistical rigor are core to your product" (Kimi K2, Feature flags)

## And against it

- "Avoid for tight budgets: LaunchDarkly/Split (free tiers cap quickly, then $10+/MAU or seats)." (Grok 4.1 Fast, Feature flags)
- "its pricing model is also complex and can be costly for smaller teams. Some users report a steep learning curve" (Mistral Small, Feature flags)
- "Split (acquired by Harness) — flags + A/B testing, but product direction has been mixed since acquisition." (DeepSeek V4 Flash, Feature flags)
- "Split.io (now Harness-integrated): Opaque pricing, rapid scaling bills, weaker CI/CD vs. peers." (Grok 4.1 Fast, Feature flags)

## Record

- Method: https://it-ai-index.com/methodology/
- Raw judge labels and full responses: https://it-ai-index.com/data/
- License: CC BY 4.0. Cite as IT AI Recommendation Index, September 2026 Edition, it-ai-index.com.