models

LiquidAI: LFM2.5-24B-A2B

Markdown · JSON

LiquidAI: LFM2.5-24B-A2B

LFM2.5-24B-A2B is LiquidAI’s text reasoning model for tool-driven agents, with 24B total and 2.45B active parameters. It reasons before answering, supports tool calls and schema-guided output, and is published with a 128,000-token context and an 8,192-token completion limit.

About the model

This description covers the model’s capabilities. Available in Ailoot inputs, outputs and options are listed for the selected method below.

LFM2.5-24B-A2B is a LiquidAI mixture-of-experts model for text-based reasoning and agents that use tools. It has 24 billion total parameters with 2.45 billion active for a token. Reasoning is part of every answer, so the task should define the goal, the evidence to use and the result the application needs.

Give an agent a bounded objective

The model suits workflows in which a language model chooses a next step, reads the resulting observation and then continues towards a deliverable. State the goal in terms the application can check: find the relevant record, compare two configurations, prepare a change proposal or produce a structured assessment.

Supply tools with clear purposes and preserve their returned observations in the task context. The model publishes tool-calling support, including selection of a tool when the interface supplies that capability. Execution belongs to the application controlling those tools. Keep a returned observation distinct from an instruction written by the model, particularly when a later step depends on the result of a search or lookup.

Retrieval for knowledge-heavy work

The publisher recommends retrieval or web search for tasks that depend on detailed external knowledge. Build the evidence step into the workflow rather than asking the model to invent facts it has not received. Identify the retrieved passages and ask the answer to connect its conclusions to those passages.

For a support workflow, this can mean finding the relevant policy and explaining how it applies to the supplied case. For a technical task, it can mean comparing the documentation and configuration before proposing a change. Keep the important retrieved content in the context so the final response can be checked without repeating every lookup.

Structured answers and sampling choices

The published model supports schema-guided structured outputs. Define a small result shape around what the application consumes, such as a selected category, an explanation, evidence references and a next action. Separate the model’s reasoning process from the fields that the next component needs to store or display.

LiquidAI lists temperature 0.1, top_k 50 and repetition_penalty 1.05 as published sampling recommendations, and the saved catalogue repeats those defaults. They provide a repeatable starting point for evaluating a specific workflow. Keep the chosen settings consistent when comparing representative tasks.

Capacity and language

The published service record lists a 128,000-token context window and an 8,192-token completion limit. Prepare a concise answer specification so the final artifact fits that output allowance. Keep relevant evidence together and leave enough output space for the requested artifact.

The maker lists a multilingual set that includes English, Russian, Chinese, Spanish, Arabic and other languages. Retain source terminology and protected names when using the model in multilingual work. A task with a fixed vocabulary benefits from an explicit glossary and the same output conventions across languages.

Capabilities

  • Text-only mixture-of-experts reasoning: 24B total, 2.45B active parameters.
  • Mandatory reasoning before the final answer.
  • Published tools, tool selection and JSON-schema structured output.
  • Retrieval-oriented workflows for knowledge-dependent tasks.
  • Published context of 128,000 tokens and completion limit of 8,192 tokens.

Use cases

  • Retrieval-backed assistant — A question, retrieved passages and a required answer layout. Result: An evidence-linked answer that can be checked against the passages.
  • Tool-driven triage — A bounded case and controlled lookup tools. Result: A structured assessment and a proposed next action.
  • Configuration comparison — Relevant configuration, documentation and comparison criteria. Result: An identified set of differences with supporting evidence.
  • Multilingual structured analysis — Source text, a glossary and consistent result fields. Result: Comparable records that retain protected terminology.

Limits

  • Context — Published service context window: 128,000 tokens.
  • Reply — Published completion limit: 8,192 tokens.
  • Reasoning — The model reasons before answering; mandatory reasoning is part of the published configuration.
  • Input/output — The published model accepts text and returns text.
Ailoot Ultra · OpenAI Chat POST /v1/chat/completions liquid/lfm-2.5-24b-a2b

liquid/lfm-2.5-24b-a2b

LiquidAI /liquid/lfm-2.5-24b-a2b

Overview

LFM2.5-24B-A2B is LiquidAI’s text reasoning model for tool-driven agents, with 24B total and 2.45B active parameters. It reasons before answering, supports tool calls and schema-guided output, and is published with a 128,000-token context and an 8,192-token completion limit.

Accepted inputs
Text
Results
Text

Available in Ailoot · inputs & methods

Ailoot Ultra · OpenAI Chat POST /v1/chat/completions
Input types
text
Output types
text
Output format
json or SSE
FieldTypeRequirementValues & limitsDefault
FieldmodelTypestringRequirementRequiredValues & limits
liquid/lfm-2.5-24b-a2b
Default—
FieldmessagesTypemessage[]RequirementRequiredValues & limits
Minimum items: 1
Default—
Fieldmessages[].roleTypestringRequirementRequiredValues & limits
Model-specific
Default—
Fieldmessages[].contentTypestringRequirementOptionalValues & limits
Applies when: Text conversation
Default—
FieldstreamTypebooleanRequirementOptionalValues & limitsDefaultfalse
Input material combinations

Text · UTF8 text

messages[].content

Method parameters and lifecycle
ParameterContract
Parametercontent_typeContractapplication/json
ParameterrequiredContractmodel, messages
Parametermaterial_typesContracttext
ParametermessagesContractmessage array with role/content shape
Parameterfile_materialContractactual UTF-8 document inserted as literal text; no file ID inferred
Parameteroptional_controlsContractstream
Parametercontrol_scopeContractText conversation; stream is a boolean and defaults to false.
Parameteroutput_modalitiesContracttext
Parameterhosted_operationsContractfalse
Parametermultimodal_input_bindingContractnull

Method aliases: /chat/completions

Lifecycle: synchronous_or_streamed

Send exact model ID with messages.

Context length is not an output cap.

A text file may supply literal text; image/audio/file-object inputs need exact binding.

Pricing

Tariff references do not add request options. Supported parameters are listed for the selected method.

Fast mode
Not specified When fast mode is selected
Flex mode
Not specified When flex mode is selected
Maximum reasoning effort
Not specified When maximum reasoning effort uses the default adjustment

Ailoot rates

Prices by configuration

Match each charge to the conditions shown. Use one rate for each set of alternative configurations.

ConfigurationRate scheduleContext tokens, more thanContext tokens, up to and includingCharge
ConfigurationBase token ratesRate scheduleBase rateContext tokens, more than—Context tokens, up to and includingNo upper boundCharge
Input · Input tokens $0.0001 / 1K tokens
Output · Output tokens $0.0014 / 1K tokens
Input · Cache read $0.00005 / 1K tokens
Input · Cache write — / 1K tokens
Input · Cache write, 1 hour — / 1K tokens
ConfigurationRequest-length tier 1Rate schedule—Context tokens, more than0Context tokens, up to and includingNo upper boundCharge
Input · Input tokens $0.0001 / 1K tokens
Output · Output tokens $0.0014 / 1K tokens
Input · Cache read $0.00005 / 1K tokens
Input · Cache write — / 1K tokens
Input · Cache write, 1 hour — / 1K tokens

Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →

Example requests

Ailoot Ultra · OpenAI Chat

curlPython SDK
set -eu
: "${AILOOT_API_KEY:?Set your Ailoot API key}"
python3 - <<'PY_REQUEST'
import json,os
from pathlib import Path
p=json.loads('{"model":"liquid/lfm-2.5-24b-a2b","messages":[{"role":"user","content":"hello"}],"stream":false}')

Path("request.json").write_text(json.dumps(p,ensure_ascii=False))
PY_REQUEST
curl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/chat/completions \
  -H "Authorization: Bearer $AILOOT_API_KEY" -H 'Content-Type: application/json' \
  --data-binary @request.json -D response-headers.txt -o result.json
python3 - <<'PY_RESULT'
import json
from pathlib import Path
r=json.loads(Path("result.json").read_text())
for x in r.get("choices",[]):
    if x.get("message",{}).get("content"): print(x["message"]["content"])
PY_RESULT
from openai import OpenAI
import os

if not os.environ.get("AILOOT_API_KEY", "").strip(): raise SystemExit("Set AILOOT_API_KEY before making a request")
client=OpenAI(base_url="https://api.ailoot.ai/v1",api_key=os.environ["AILOOT_API_KEY"],max_retries=0)
result=client.chat.completions.create(model="liquid/lfm-2.5-24b-a2b", messages=[{"role":"user","content":"hello"}], stream=False)
import base64, json
from pathlib import Path

def save_result(data, stem="result"):
    # Keep every field, including usage, IDs, reasoning and tool requests.
    Path(stem + ".json").write_text(json.dumps(data, ensure_ascii=False, indent=2), encoding="utf-8")
    print(json.dumps(data, ensure_ascii=False, indent=2))
    def save_blob(encoded, suffix, label):
        raw = base64.b64decode(encoded, altchars=b"-_", validate=True)
        ext = "png" if raw.startswith(b"\x89PNG\r\n\x1a\n") else "jpg" if raw.startswith(b"\xff\xd8") else "webp" if raw[:4] == b"RIFF" and raw[8:12] == b"WEBP" else suffix
        file = Path(stem + "-" + label + "." + ext)
        file.write_bytes(raw)
        print("Saved", file)
    for index, block in enumerate(data.get("content") or []):
        print("Content", index, json.dumps(block, ensure_ascii=False))
    for choice in data.get("choices") or []:
        print("Choice", choice.get("index"), json.dumps(choice.get("message", choice.get("delta", {})), ensure_ascii=False))
    for index, item in enumerate(data.get("output") or []):
        print("Output", index, json.dumps(item, ensure_ascii=False))
        if item.get("type") == "image_generation_call" and item.get("result"):
            save_blob(item["result"], item.get("output_format") or "bin", "output-" + str(index))
    for ci, candidate in enumerate(data.get("candidates") or []):
        for pi, part in enumerate((candidate.get("content") or {}).get("parts") or []):
            print("Candidate/part", ci, pi, json.dumps(part, ensure_ascii=False))
            blob = part.get("inlineData") or part.get("inline_data")
            if blob and blob.get("data"):
                mime = blob.get("mimeType") or blob.get("mime_type")
                suffix = {"image/png":"png", "image/jpeg":"jpg", "image/webp":"webp", "audio/wav":"wav", "audio/mpeg":"mp3", "video/mp4":"mp4"}.get(mime, "bin")
                save_blob(blob["data"], suffix, "candidate-" + str(ci) + "-part-" + str(pi))

save_result(result.to_dict(mode="json"))

Receive and save results

Ailoot Ultra · OpenAI Chat

json or SSE

Keep all completed results

Keep every complete returned response, file and content part.

Keep every choice and its content/tool calls/reasoning. In SSE assemble deltas by choice/tool index, retain the final finish reason and usage.

If the original response needs checking

Keep request/response IDs and headers. No generic text job retrieval URL is supplied.

Response fields
FieldType
Fieldchoices[].indexTypeinteger
Fieldchoices[].message.contentTypestring|null
Fieldchoices[].message.tool_callsTypetool_call[]
Fieldchoices[].message.reasoning_contentTypestring
Fieldchoices[].message.reasoning_detailsTypeobject|array
Fieldchoices[].finish_reasonTypestring|null
FieldusageTypeobject
Errors and next steps
ResponseConditionNext step
Response400 · invalid_request_errorConditionInvalid required fields, material representation or parameter valueNext stepCorrect the indicated fields before a new creation request.

Aliases

Model aliases

Ailoot Ultra liquid/lfm-2.5-24b-a2b

Method path aliases

/chat/completions

Related caller IDs

liquid/lfm-2.5-24b-a2b

Ailoot Ultra · Anthropic Messages POST /v1/messages liquid/lfm-2.5-24b-a2b

liquid/lfm-2.5-24b-a2b

LiquidAI /liquid/lfm-2.5-24b-a2b

Overview

LFM2.5-24B-A2B is LiquidAI’s text reasoning model for tool-driven agents, with 24B total and 2.45B active parameters. It reasons before answering, supports tool calls and schema-guided output, and is published with a 128,000-token context and an 8,192-token completion limit.

Accepted inputs
Text
Results
json · sse

Available in Ailoot · inputs & methods

Ailoot Ultra · Anthropic Messages POST /v1/messages
Input types
text
Output types
json, sse
Output format
JSON or SSE
FieldTypeRequirementValues & limitsDefault
FieldmodelTypestringRequirementRequiredValues & limits
liquid/lfm-2.5-24b-a2b
Default—
FieldmessagesTypemessage[]RequirementRequiredValues & limitsDefault—
Fieldmessages[].roleTypestringRequirementRequiredValues & limits
user, assistant
Default—
Fieldmessages[].contentTypestringRequirementRequiredValues & limitsDefault—
Fieldmax_tokensTypeintegerRequirementRequiredValues & limitsDefault—
FieldstreamTypebooleanRequirementOptionalValues & limitsDefaultfalse
Input material combinations

Text · UTF-8 text

messages[].content string

Method parameters and lifecycle
ParameterContract
Parametercontent_typeContractapplication/json
ParameterrequiredContractmodel, messages, max_tokens
Parameteroptional_controlsContractstream
Parametermaterial_typesContracttext
Parametertext_file_materialContractRead the actual UTF-8 document and put its text in the indicated input or message content string.

Lifecycle: synchronous_or_streamed

Pricing

Tariff references do not add request options. Supported parameters are listed for the selected method.

Fast mode
Not specified When fast mode is selected
Flex mode
Not specified When flex mode is selected
Maximum reasoning effort
Not specified When maximum reasoning effort uses the default adjustment

Ailoot rates

Prices by configuration

Match each charge to the conditions shown. Use one rate for each set of alternative configurations.

ConfigurationRate scheduleContext tokens, more thanContext tokens, up to and includingCharge
ConfigurationBase token ratesRate scheduleBase rateContext tokens, more than—Context tokens, up to and includingNo upper boundCharge
Input · Input tokens $0.0001 / 1K tokens
Output · Output tokens $0.0014 / 1K tokens
Input · Cache read $0.00005 / 1K tokens
Input · Cache write — / 1K tokens
Input · Cache write, 1 hour — / 1K tokens
ConfigurationRequest-length tier 1Rate schedule—Context tokens, more than0Context tokens, up to and includingNo upper boundCharge
Input · Input tokens $0.0001 / 1K tokens
Output · Output tokens $0.0014 / 1K tokens
Input · Cache read $0.00005 / 1K tokens
Input · Cache write — / 1K tokens
Input · Cache write, 1 hour — / 1K tokens

Ailoot reserves an estimated amount before a request. After the result, it settles the final charge from measured usage. Streaming settlement uses the final usage record. How reservation and settlement work →

Example requests

Ailoot Ultra · Anthropic Messages

Complete minimal JSON bodycurl · use your UTF-8 documentPython · save the complete result
{
  "model": "liquid/lfm-2.5-24b-a2b",
  "max_tokens": 128,
  "messages": [
    {
      "role": "user",
      "content": "Summarize the incident report."
    }
  ],
  "stream": false
}
set -eu
: "${AILOOT_API_KEY:?Set your Ailoot API key}"
: "${AILOOT_TEXT_FILE:?Set your actual UTF-8 document}"
export AILOOT_TEXT_FILE
python3 - <<'PY_REQUEST'
import json, os
from pathlib import Path

source_file = Path(os.environ['AILOOT_TEXT_FILE'])
if not source_file.is_file():
    raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')
text = source_file.read_text(encoding='utf-8')
if not text.strip():
    raise SystemExit('Choose a document containing text')
payload = {'model': 'liquid/lfm-2.5-24b-a2b', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}
payload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\n\n' + text
Path('request.json').write_text(json.dumps(payload, ensure_ascii=False), encoding='utf-8')
PY_REQUEST
curl --fail-with-body --silent --show-error https://api.ailoot.ai/v1/messages \
  -H "x-api-key: $AILOOT_API_KEY" -H 'anthropic-version: 2023-06-01' -H 'Content-Type: application/json' \
  --data-binary @request.json -D response-headers.txt -o result.json
python3 - <<'PY_RESULT'
import json
from pathlib import Path

result = json.loads(Path('result.json').read_text(encoding='utf-8'))
print(json.dumps(result, ensure_ascii=False, indent=2))
PY_RESULT
import json, os
from pathlib import Path

source_file = Path(os.environ['AILOOT_TEXT_FILE'])
if not source_file.is_file():
    raise SystemExit('Set AILOOT_TEXT_FILE to your actual UTF-8 document')
text = source_file.read_text(encoding='utf-8')
if not text.strip():
    raise SystemExit('Choose a document containing text')
payload = {'model': 'liquid/lfm-2.5-24b-a2b', 'max_tokens': 128, 'messages': [{'role': 'user', 'content': 'Summarize the incident report.'}]}
payload['messages'][0]['content'] = 'Summarize this material and quote supporting source passages.\n\n' + text
from urllib.request import Request, urlopen
from urllib.error import HTTPError

api_key = os.environ['AILOOT_API_KEY']
if not api_key.strip():
    raise SystemExit('Set AILOOT_API_KEY before making a request')
headers = {'x-api-key': 'PLACEHOLDER_KEY', 'anthropic-version': '2023-06-01', 'Content-Type': 'application/json'}
headers['x-api-key'] = api_key
request = Request('https://api.ailoot.ai/v1/messages', data=json.dumps(payload, ensure_ascii=False).encode('utf-8'), headers=headers, method='POST')
try:
    with urlopen(request, timeout=120) as response:
        raw = response.read()
        Path('response-headers.txt').write_text(str(response.headers), encoding='utf-8')
except HTTPError as error:
    Path('result.json').write_bytes(error.read())
    Path('response-headers.txt').write_text(str(error.headers), encoding='utf-8')
    raise
Path('result.json').write_bytes(raw)
import json
from pathlib import Path

result = json.loads(Path('result.json').read_text(encoding='utf-8'))
print(json.dumps(result, ensure_ascii=False, indent=2))

Receive and save results

Ailoot Ultra · Anthropic Messages

JSON or SSE

Keep all completed results

Keep every complete returned response, file and content part.

Keep all content blocks, thinking signatures and actual tool-use IDs. Assemble all SSE block deltas; a first text block is not the complete output.

If the original response needs checking

Keep request/response IDs and headers. No generic text job retrieval URL is supplied.

Response fields
FieldType
FieldidTypestring
FieldmodelTypestring
Fieldcontent[]Typetext|thinking|tool_use|other content block
Fieldstop_reasonTypestring|null
FieldusageTypeobject
Errors and next steps
ResponseConditionNext step
Response400 · invalid_request_errorConditionInvalid required fields, material representation or parameter valueNext stepCorrect the indicated fields before a new creation request.

Aliases

Model aliases

Ailoot Ultra liquid/lfm-2.5-24b-a2b

Method path aliases

/chat/completions

Related caller IDs

liquid/lfm-2.5-24b-a2b

FAQ · LiquidAI: LFM2.5-24B-A2B
What inputs and results are available?

POST /v1/chat/completions · liquid/lfm-2.5-24b-a2b: text → text

POST /v1/messages · liquid/lfm-2.5-24b-a2b: text → json, sse

Which fields does the request use?

POST /v1/chat/completions · liquid/lfm-2.5-24b-a2b: model, messages, stream

POST /v1/messages · liquid/lfm-2.5-24b-a2b: model, messages, max_tokens, stream

What limits apply?

POST /v1/chat/completions · liquid/lfm-2.5-24b-a2b: messages.min_items: 1

More details

Connect Ailoot to Claude Code, OpenCode, Hermes or your own app.

The company and its project portfolio

The company and its project portfolio: millions of customers; monthly sales in the billions.

Buy tokens at a discount.

Subscription discounts for regular use. Promo codes offer up to 50% off token purchases.

Subscription
Choose a discounted plan for your regular requests.
Promo code: up to 50% off
Enter your code at checkout. The price updates to reflect the current promotion.
Spending in your history
The model, tokens and amount are recorded for each request.

Choose a subscription · View model prices

Connect the API to a product you sell.

Use Ailoot in SaaS, apps and bots under its official service offer and licence agreement. Request a reseller licence for resale.

Commercial application
Add text, image, video and audio generation to paid features in your service.
Reseller licence
Send a description of your project and how you provide access. Request written terms for use and resale.
Documents on the website
The service offer, licence agreement and data processing policy are available before you connect.

Request a reseller licence · Service offer and licence agreement