NOUMENOpen app
NOUMEN

liquid · Open weights

LFM2.5 2.6B

Added Aug 12, 2026Open weights

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Liquid AI's compact 2.6B reasoning model for agent workflows, data extraction, RAG, and long-context processing. It supports tool calling and structured output, but Liquid advises against using it for agentic coding. Warning: prompts and responses may be logged and used for model training or service improvement; do not send sensitive data.

  • Reasoning
  • Tools
  • Structured output
Input
Text
Output
Text
Reasoning effort
minimal · low · medium · high · xhigh
Discovery tier
Open weights

Pricing

USD per 1M tokens
Input
$0.10
Output
$0.20

Reference prices as published by the model provider. API example

Specifications

Context window
128K
Max output
32.8K
Average output (7 days)
459 tokens
Longer than 45% of 465 models

Side by side

LFM2.5 2.6B compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

LFM2.5 2.6BliquidThis modelQwen 3 235b A22BQwenRemove
ProviderliquidQwen
Discovery tierOpen weightsOpen weights
AddedAug 12, 2026Apr 29, 2025
Limits
Context window128K262.14KBest
Max output32.77KBest16.38K
Average output459 tokens616 tokens
Pricing, USD per 1M tokens
Input$0.10$0.0875Best
Output$0.20Best$0.35
Cache read—$0.0175
Performance
Request success (24h)—91.3%
Benchmark indexes
Intelligence Index8.4Best8.3
Coding Index7.7—
Math Index—23.7
Capabilities
ReasoningYesNo
ToolsYesYes
Structured outputYesYes
PDF inputNoYes
Reasoning effortminimal, low, medium, high, xhigh—

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

  • LFM2.5 2.6B
  • Qwen 3 235b A22B

Intelligence Index

  • LFM2.5 2.6B8.4
  • Qwen 3 235b A22B8.3

LFM2.5 2.6B: Better than 13% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • LFM2.5 2.6B: 7.7
    • Qwen 3 235b A22B: —

  • Math Index

    Composite of the math evaluations

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 23.7

Document reasoning

  • AA-LCRv1.1

    Reasoning across long documents

    • LFM2.5 2.6B: 5.7%Top
    • Qwen 3 235b A22B: 0.0%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • LFM2.5 2.6B: 6.2%Top
    • Qwen 3 235b A22B: 4.2%

  • GPQA Diamond

    Graduate-level science questions

    • LFM2.5 2.6B: 55.8%
    • Qwen 3 235b A22B: 61.3%Top

  • MMLU-Pro

    Multi-subject knowledge and reasoning

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 76.2%

  • IFBench

    Following precise output instructions

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 36.6%

Coding

  • SciCode

    Scientific computing in Python

    • LFM2.5 2.6B: 14.4%
    • Qwen 3 235b A22B: —

  • LiveCodeBench

    Recent competitive programming problems

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 34.3%

Math

  • AIME 2025

    Competition mathematics, 2025 problems

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 23.7%

  • AIME

    Competition mathematics

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 32.7%

  • MATH-500

    Competition math problems

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 90.2%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 6.1%

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • LFM2.5 2.6B: —
    • Qwen 3 235b A22B: 27.2%

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • LFM2.5 2.6B: 5.7%Top
    • Qwen 3 235b A22B: 0.0%

Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart
curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "liquid/lfm-2.5-2.6b",
    "reasoning_effort": "medium",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'
import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="liquid/lfm-2.5-2.6b",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

liquid

Other liquid models, starting with the same family.