How to Use LFM3-1.7B in Muse
LFM3-1.7B suits focused text tasks, tool calling, and structured output with a 256K-token context window. Liquid AI advises against using it for agentic coding or complex multistep reasoning.
Prompts to try with LFM3-1.7B in Muse
Copy an example into Muse, or ask in your own words.
Use LFM3-1.7B to classify the messages I provide as cancellation, delivery question, product question, or uncertain. Return one category and a supporting quote for each message.
Use LFM3-1.7B to extract the customer name, order number, requested action, and stated deadline from the text I provide. Return JSON and use null for missing details.
Use LFM3-1.7B to answer my question using only the reference excerpts I provide. Keep the answer under 150 words, quote the supporting passage, and identify anything the excerpts do not establish.
Use LFM3-1.7B to label the intent of the English and Spanish customer messages I provide. Use billing, delivery, cancellation, or uncertain, and preserve the original text beside each label.
Use LFM3-1.7B with temperature set to 0.1 to extract commitments from the conversation text I provide. List the person, promised action, and stated deadline; mark missing details as unknown.
How to set up LFM3-1.7B in Muse
Pick the way that suits you. Each takes about two minutes.
Use the Muse connector
- Create an API key
Create one under API keys in the Spiral console.
- Send Muse this message
Muse reads Spiral's skill and asks for an API key.
set up https://spiral.ai/SKILL.md
- Give Muse the key
Paste the key when Muse asks for it, never in the chat itself.
Use the Spiral skill
- Send Muse this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run LFM3-1.7B
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openrouter", "model": "liquid/lfm-3-1.7b", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Muse finds LFM3-1.7B on Spiral, shows the price and runs it.
LFM3-1.7B providers and prices
Spiral runs LFM3-1.7B through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.001069 per run at the default inputs. $0.104 per 1M input tokens, $0.26 per 1M output tokens.
Settings you can ask for
Ask Muse for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 8192 | The most output tokens to generate. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 |
Every API field of LiquidAI: LFM3-1.7B is on its tool page: open its card above.
What can you do with LFM3-1.7B in Muse?
Turn messages and reference excerpts into labeled records or concise answers with LFM3-1.7B. Supply the relevant text and define the result you need.
Support request triage
Ask LFM3-1.7B to label each request using your queue definitions and quote the words that justify the label.
Structured record extraction
Turn order messages or intake notes into JSON with LFM3-1.7B. Specify the required fields and use null for missing information.
Answers from supplied reference text
Give LFM3-1.7B relevant help-center excerpts and ask for a short answer with supporting passages and any information still missing.
Tool selection for routine lookups
Ask LFM3-1.7B to select a lookup tool for an order or inventory question, then explain the returned result.
Issue labeling
Have LFM3-1.7B classify issue descriptions against a fixed taxonomy, returning an uncertainty label when the text does not establish a category.
Multilingual message routing
Use LFM3-1.7B to label intent in messages written in its supported languages, keeping the original wording beside each classification for review.
Log detail extraction
Ask LFM3-1.7B to extract timestamps, error codes, and affected services from log text into records without inferring a root cause.
More on LFM3-1.7B and Muse
Overview
LFM3-1.7B in other agents
Questions about LFM3-1.7B in Muse
Can Muse use LFM3-1.7B?
Yes. Add Spiral to Muse as an MCP server, and Muse can find and run LFM3-1.7B like any built-in tool.
How much does LFM3-1.7B cost in Muse?
Up to $0.001069 per run at the default inputs. $0.104 per 1M input tokens, $0.26 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Muse plan is separate.
Do I need an OpenRouter account to use LFM3-1.7B in Muse?
You need no OpenRouter account. Muse uses one Spiral API key for every tool, and Spiral runs LFM3-1.7B on its own OpenRouter account.
How long does it take to set up LFM3-1.7B in Muse?
About two minutes: create an API key, then send Muse this message, then give Muse the key.
Does Muse wait while LFM3-1.7B runs?
LFM3-1.7B runs in the background. Muse starts the run, checks on it and gives you the text, json when it's ready.
What is LFM3-1.7B best suited to?
Use it for focused tasks such as extracting records, labeling requests, producing structured output, and selecting tools. Give it a clear task and a specific result to return.
How much text can LFM3-1.7B handle?
Its context window holds 256K tokens across input and output combined. Leave room for the response when supplying long text.
Which languages does LFM3-1.7B support?
Liquid AI lists English, Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Spanish, Vietnamese, Thai, Indonesian, Hindi, Russian, and Polish. Test your own terminology and examples before using it for a language-specific workflow.
Is LFM3-1.7B suitable for autonomous coding?
Liquid AI advises against agentic coding and complex multistep reasoning. In a coding assistant, keep requests bounded, such as extracting details from logs or classifying issues against supplied rules.
Which sampling settings should I start with?
Liquid AI's generation example uses temperature 0.1 and top-k 50. Start there for focused text tasks, then compare results against examples you have already checked.
Start using LFM3-1.7B in Muse
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Muse:
set up https://spiral.ai/SKILL.md


