Pi (Inflection-2.5)

Conversational

Inflection's Pi assistant model (v2.5). After Microsoft's 2024 licensing deal absorbed most of the Inflection team, Pi continues as an enterprise emotional-intelligence assistant.

Pi (Inflection-2.5) is a frontier AI model from Inflection AI. It costs $2.500 per million input tokens and $10.000 per million output tokens (blended $4.750/M), with a 32,000-token context window.

INPUT
$2.500/M
per million input tokens
OUTPUT
$10.000/M
per million output tokens
BLENDED 70/30
$4.750/M
unchanged since 3 May
CONTEXT
32,000
tokens
What it is good at
  • Conversational warmth
  • Long memory
  • Tuned for assistant use
Typical use cases
  • Consumer companion / coaching apps
  • Emotional-intelligence enterprise chat

Benchmarks

vs. best public score
MMLU73%
Multitask academic knowledge across 57 subjects.
LMArena Elo1163 Elo
Crowd-sourced head-to-head preference Elo rating.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

Frequently asked questions

How much does Pi (Inflection-2.5) cost?

Pi (Inflection-2.5) costs $2.500 per million input tokens and $10.000 per million output tokens, for a blended reference rate of $4.750 per million tokens.

What is Pi (Inflection-2.5)'s context window?

Pi (Inflection-2.5) supports up to 32,000 tokens of context in a single request.

What is Pi (Inflection-2.5) best for?

Pi (Inflection-2.5) is well suited to Conversational warmth, Long memory and Tuned for assistant use.

Who makes Pi (Inflection-2.5)?

Pi (Inflection-2.5) is developed and served by Inflection AI. It was released in Mar 2024.

Terms used on this page