Skip to content

Model Library

Liquid AIReleased February 25, 2026

LFM2 24B A2B

Liquid AI · Released February 25, 2026

View public research Browse the library

Intelligence score

5

At a glance

Intelligence
5
Speed
114.4tokens/sec
Latency
570mstime to first token
Input price
$0.03per 1M tokens
Output price
$0.12per 1M tokens
Context window
33Ktokens
Parameters
23.8billions

Summary

Comparison summary

LFM2 24B A2B scores 5 (estimated) on the Artificial Analysis Intelligence Index, placing it below average among other open weight non-reasoning models of similar size (median: 6). LFM2 24B A2B generates output at 114.4 tokens per second (based on the median across providers serving the model), which is above average compared to other open weight non-reasoning models of similar size (median: 101.8 t/s). LFM2 24B A2B costs $0.03 per 1M input tokens (very competitive, median: $0.15) and $0.12 per 1M output tokens (very competitive, median: $0.32), based on the median across providers serving the model.

Specifications

Technical specifications

Model nameLFM2 24B A2B
ProviderLiquid AI
Release dateFebruary 25, 2026
Intelligence score5
Speed114.4 tokens/sec
Latency (TTFT)0.57s
Input price$0.03 / 1M tokens
Output price$0.12 / 1M tokens
Context window33,000 tokens
Parameters23.8B
ReasoningNo
Input modalitiestext
Output modalities—
Open weightsYes
Licenselfm 1
API availability—
SourceModel Intelligence legacy profile: lfm2-24b-a2b

Pricing

Token prices

Price per 1M tokens

LFM2 24B A2B, as listed in the catalog record

Input $0.03 and output $0.12 per 1M tokens. Missing prices are hatched and marked "Not published".
  1. Input$0.03
  2. Output$0.12

Model Intelligence legacy profile: lfm2-24b-a2b

FAQ · 20

Frequently asked questions

LFM2 24B A2B was released on February 25, 2026.

LFM2 24B A2B was created by Liquid AI.

LFM2 24B A2B scores 5 (estimated) on the Artificial Analysis Intelligence Index, placing it below average among other open weight non-reasoning models of similar size (median: 6).

LFM2 24B A2B generates output at 114.4 tokens per second (based on the median across providers serving the model), which is above average compared to other open weight non-reasoning models of similar size (median: 101.8 t/s).

LFM2 24B A2B has a time to first token (TTFT) of 0.57s (based on the median across providers serving the model), which is very competitive compared to other open weight non-reasoning models of similar size (median: 1.54s).

LFM2 24B A2B costs $0.03 per 1M input tokens (very competitive, median: $0.15) and $0.12 per 1M output tokens (very competitive, median: $0.32), based on the median across providers serving the model.

LFM2 24B A2B costs $0.03 per 1M input tokens and $0.12 per 1M output tokens (based on the median across providers serving the model). For a blended rate (7:2:1 cache hit/input/output ratio), this is $0.04 per 1M tokens. Pricing may vary by provider.

No, LFM2 24B A2B is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.

LFM2 24B A2B supports text input.

LFM2 24B A2B supports text output.

No, LFM2 24B A2B does not support image input. It can only process text.

No, LFM2 24B A2B is not multimodal. It only supports text input.

LFM2 24B A2B has a context window of 33k tokens. This determines how much text and conversation history the model can process in a single request.

Yes, LFM2 24B A2B is open weights. The model weights are publicly available and can be downloaded for self-hosting.

LFM2 24B A2B has 23.8 billion parameters (2.3 billion active).

LFM2 24B A2B is a Mixture of Experts (MoE) model with 23.8 billion total parameters, but only 2.3 billion active parameters are used during inference.

LFM2 24B A2B is released under the lfm 1.0 license. This license allows commercial use.

LFM2 24B A2B achieves a score of 5 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.

Yes, LFM2 24B A2B is available via API through 1 provider.

LFM2 24B A2B is available through 1 API provider.