Inflection-2 Models by Inflection
Last refreshed 2026-05-19. Next refresh: weekly.
Details
Links
WebsiteAbout
Inflection-2, developed by Inflection AI, is an advanced large language model (LLM) hailed for its remarkable capabilities, ranking second globally, just behind OpenAI's GPT-4. Created by Inflection AI, founded by Mustafa Suleyman and Reid Hoffman, this model builds on its predecessor, Inflection-1, offering enhanced factual knowledge, reasoning abilities, and stylistic control. Notably outperforming competitors like Google's PaLM 2, Inflection-2 is built with efficiency in mind, powered by 5,000 NVIDIA H100 GPUs and achieving a processing milestone of roughly 10²⁵ FLOPs. It's slated to drive the company's AI-driven chatbot, Pi, and represents a step forward in their ambition to craft significantly larger models with a 22,000 GPU cluster, emphasizing ethical development and responsible scaling. Inflection's commitment to ethical AI is reinforced by signing the White House's AI principles pact 1.
Decision facts
- Best fit
- chatbot and role-playing use cases
- Capability starting point
- Inflection-2
- Lowest tracked input
- Not tracked
- Closest related family
- Inflection 3
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Inflection-2 | Use when provider availability and model metadata match the workload. | 2023-11 | — | Current |
Release Timeline
1 release groupSpecifications(1 models)
| Model | Released |
|---|---|
| Inflection-2 | 2023-11 |
