- Size
- 1.7B parameterscounted from the safetensors header on the Hub
- Licence
- custom, read it, LFM Open License v1.0Our reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
- On the Hub since
- 23 Oct 2026the date the repo was created
- Context length
- 256k tokensfrom the model's config
- Fine-tune
- of LiquidAI/LFM3-1.7B-BaseLfm3ForCausalLM
- On Hugging Face
- LiquidAI/LFM3-1.7B120k downloads in the last 30 days, 2.0k likes
Scores
No independent scores for this model in Epoch AI's data. A score is only shown here when Epoch's name for a model matches this repo with certainty; an uncertain match is left out rather than guessed.
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.10 GB, worked out from the model's config (LiquidAI/LFM3-1.7B), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
LiquidAI/LFM3-1.7B-GGUF (the maker's own conversion), 405k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| Q4_0 | 0.98 GBfile size | 1.6 GB0.98 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q4_0Ollama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q4_0Download hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-QAD-Q4_0.gguf" --local-dir LFM3-1.7B-GGUF |
| Q4_K_M | 1.0 GBfile size | 1.6 GB1.0 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q4_K_MOllama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q4_K_MDownload hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-Q4_K_M.gguf" --local-dir LFM3-1.7B-GGUF |
| Q5_K_M | 1.2 GBfile size | 1.8 GB1.2 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q5_K_MOllama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q5_K_MDownload hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-Q5_K_M.gguf" --local-dir LFM3-1.7B-GGUF |
| Q6_K | 1.3 GBfile size | 1.9 GB1.3 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q6_KOllama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q6_KDownload hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-Q6_K.gguf" --local-dir LFM3-1.7B-GGUF |
| Q8_0 | 1.7 GBfile size | 2.3 GB1.7 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q8_0Ollama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q8_0Download hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-Q8_0.gguf" --local-dir LFM3-1.7B-GGUF |
| BF16 | 3.2 GBfile size | 3.8 GB3.2 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:BF16Ollama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:BF16Download hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-BF16.gguf" --local-dir LFM3-1.7B-GGUF |
| F16 | 3.2 GBfile size | 3.8 GB3.2 + 0.10 + 0.50 | Commandsllama.cpp llama-server -hf LiquidAI/LFM3-1.7B-GGUF:F16Ollama ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:F16Download hf download LiquidAI/LFM3-1.7B-GGUF --include "LFM3-1.7B-F16.gguf" --local-dir LFM3-1.7B-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX 4bitLiquidAI/LFM3-1.7B-MLX-4bit | 0.91 GBfile size | 1.5 GB0.91 + 0.10 + 0.50 | Download page |
| MLX 8bitLiquidAI/LFM3-1.7B-MLX-8bit | 1.7 GBfile size | 2.3 GB1.7 + 0.10 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 3.2 GBfile size | 3.8 GB3.2 + 0.10 + 0.50 |
hf download LiquidAI/LFM3-1.7B --local-dir LFM3-1.7BOther conversions (7; sizes not read)
- LiquidAI/LFM3-1.7B-NVFP4 NVFP4, by the maker, 5.8k downloads
- LiquidAI/LFM3-1.7B-FP8 FP8, by the maker, 3.4k downloads
- LiquidAI/LFM3-1.7B-ONNX Other, by the maker, 3.1k downloads
- litert-community/LFM3-1.7B-LiteRT Other, 3.7k downloads
- onnx-community/LFM3-1.7B-ONNX Other, 2.3k downloads
- infosave/LFM3-1.7B-cmf Other, 1.4k downloads
- BennyDaBall/LFM3-1.7B-NVFP4 GPTQ, nvfp4, 733 downloads
The Hub tags 100 or more quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
1 provider serves this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.
| Provider | In | Out | Context | Precision | Tokens/s | First token | Uptime 24h | Tools | JSON |
|---|---|---|---|---|---|---|---|---|---|
| Liquidthe maker | $0.08 | $0.20 | 256k | fp8 | 119 | 0.6 s | 100.0% | yes | yes |
Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.
curl https://router.huggingface.co/v1/chat/completions \
-H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \
-d '{"model": "LiquidAI/LFM3-1.7B:liquid", "messages": [{"role": "user", "content": "Hello"}]}'Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Fine-tunes of this model
Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.
- orcarouter/LFM3-1.7B-abliterated1.7B795Unreviewed
- Peymandc/LFM3-1.7B-Heretic-Uncensored1.7B360Unreviewed
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.