Liquid AILanguage modelOpen weights

LFM3-1.7B

LFM3-1.7B has no overall index value yet because there are too few independent benchmark results.

Summary

LFM3-1.7B is a dense text model from Liquid AI with 1.73 billion parameters and 262,144 context tokens. Its 24 layers combine short convolutions with local and global attention. Liquid AI post-trained it for local agents, tool calling, structured responses, and task-specific adaptation.

The publisher card lists 29 text benchmarks and their unweighted mean, plus edge measurements on three devices and a B300 throughput claim. All values apply to the documented configurations and have not been independently reproduced.

All benchmark results

Every published result with its source, test setup and comparison models.

Show all 40 published benchmark results

Measurements without matching peer values

These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.

LiveClaw (Show measurement, test conditions, and source)

Source value: 45.6
Score: 45.6
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

Claw-Eval General (EN), average (Show measurement, test conditions, and source)

Source value: 68.64
Score: 68.64
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

Claw-Eval General (EN), pass^3 (Show measurement, test conditions, and source)

Source value: 51.4
Score: 51.4
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

BrowseComp-Plus (OpenClaw subset) (Show measurement, test conditions, and source)

Source value: 37
Score: 37
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

WebArena-Verified (Show measurement, test conditions, and source)

Source value: 38.42
Score: 38.42
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

EnterpriseOps-Gym (Show measurement, test conditions, and source)

Source value: 28.2
Score: 28.2
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

AutomationBench (public) (Show measurement, test conditions, and source)

Source value: 41.29
Score: 41.29
Metric: percent
Unit: %
Category: Agents
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

BFCL V4 (Show measurement, test conditions, and source)

Source value: 71.39
Score: 71.39
Metric: percent
Unit: %
Category: Tool use
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

ToolSandbox (Show measurement, test conditions, and source)

Source value: 80.36
Score: 80.36
Metric: percent
Unit: %
Category: Tool use
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

MCP-Atlas (Show measurement, test conditions, and source)

Source value: 32.57
Score: 32.57
Metric: percent
Unit: %
Category: Tool use
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

τ³-bench (4 domains), average (Show measurement, test conditions, and source)

Source value: 68.01
Score: 68.01
Metric: percent
Unit: %
Category: Tool use
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

SWE-bench Verified (OpenCode) (Show measurement, test conditions, and source)

Source value: 43.4
Score: 43.4
Metric: percent
Unit: %
Category: Agentic coding
Direction: Higher is better
Source type: vendor-reported
Evaluator: LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7
Status: active
Retrieved at: 2026-10-29
Methodology: Vendor-reported result. vLLM BF16; full 262,144-token context; unless noted, temperature 0.6 and top_p 0.95; output limit is the context length.
Context: Manufacturer-reported; these sources do not provide an independent reproduction of this result.

Source LiquidAI/LFM3-1.7B official model card at revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7

Page 1 of 4

All specifications

Architecture

Architecture: Dense tri-hybrid architecture with 24 layers: 13 gated short convolutions, 5 sliding-window attention layers with a 256-token window, and 6 global attention layers. The convolution kernel size is 7.; Source: Architecture

Input and output

Input modalities: Text; Source: Input modalities
Output modalities: Text; Source: Output modalities

Licence and access

Access: Open weights, local deployment, and the API and Playground linked by the publisher; Source: Access
License: LFM Open License v1.0. See the license text for its terms.; Source: License

Further details

Hugging Face model ID: LiquidAI/LFM3-1.7B; Source: Hugging Face model ID
Exact parameter count: 1,728,801,536 in Hugging Face model metadata.; Source: Exact parameter count
Vocabulary and languages: 128,000 tokens; 16 listed languages: English, Arabic, Chinese, French, German, Italian, Japanese, Korean, Portuguese, Spanish, Vietnamese, Thai, Indonesian, Hindi, Russian, and Polish.; Source: Vocabulary and languages
Generation defaults: The model card lists temperature=0.1, top_k=50, and repetition_penalty=1.1. It recommends separate values for math and science.; Source: Generation defaults
Fine-tuning: The publisher provides SFT, DPO, GRPO, and CPT notebooks. These are adaptation recipes, not additional profiles.; Source: Fine-tuning

Frequently asked questions about LFM3-1.7B

How large is the context window of LFM3-1.7B?

262,144 tokens. That is how much text LFM3-1.7B reads at once, about 390 A4 pages.

Page count as a rule of thumb from Artificial Analysis

Context windowTokens the model reads at once · higher is better

LFM3-1.7BRank 68 of 100262K

Kimi K2.6262K

Qwen 3.8 27B262K

Kimi K2.7 Code262K

Kimi K2.5262K

Kimi K2 Thinking262K

Ling 3.0 Flash262K

Can I run LFM3-1.7B locally?

Yes, the weights are open.

Stated by Liquid AI

Open weights: yes
License: LFM Open License v1.0. See the license text for its terms.
Parameters: 1.7B
active per token: n/a

Sources and data date

Show all 10 sources

Benchmark sources at a glance

Which independent evaluators tested this model. This is coverage, not a quality ranking.

No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.

Without vendor claims and historical or disputed results; those stay labelled in the single values.

Type	Evidence and data date
Input	Documented modalities. Retrieved October 29, 2026. Source: LiquidAI/LFM3-1.7B model card at pinned revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7 · LiquidAI/LFM3-1.7B model card at pinned revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7 (retrieved October 29, 2026) · LFM3-1.7B model card · Editorial description reviewed October 29, 2026
Output	Documented modalities. Retrieved October 29, 2026. Source: LiquidAI/LFM3-1.7B model card at pinned revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7 · LiquidAI/LFM3-1.7B model card at pinned revision ac2586a34f1a1c03329dbdcdefabd8b6478ddaa7 (retrieved October 29, 2026) · LFM3-1.7B model card · Editorial description reviewed October 29, 2026
Research status	Research date October 29, 2026. 63 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not published	Throughput chart raw points Note The chart does not publish a machine-readable numeric table for the other concurrency levels.
Not found in the inspected sources	Independent evaluation Note The reviewed manufacturer sources do not link an independent reproduction of these benchmarks.
Not found in the inspected sources	Technical report or research paper Note Reviewed publisher sources do not link a separate technical report or research paper for LFM3-1.7B.
Not published	Sample counts Note The benchmark table does not state per-suite sample counts.
Not found in the inspected sources	Vals rows in captured payloads Note Nine complete Vals payloads dated 28 October were checked; no exact model ID or unambiguous model label. Limited to these saved benchmarks.
Unresolved	Uncaptured Arena Code/Speech leaderboards Note Text, Agent including Code/Chat/Work, and twelve WebDev views were captured. A separate general Code and Arena Speech leaderboard were not saved as complete tables.
Unresolved	Artificial Analysis catalog reconciliation Note The full current AA model and benchmark catalog was not re-reconciled for this profile in this sidecar. This sidecar makes no AA absence claim; native AA imports remain authoritative.
Additional source	Liquid AI LFM3 release announcement (retrieved October 29, 2026) · Liquid AI LFM3 release announcement · Editorial description reviewed October 29, 2026
Additional source	Liquid AI model documentation (retrieved October 29, 2026)
Additional source	Liquid AI LFM3 Hugging Face collection (retrieved October 29, 2026) · LFM3 collection · Editorial description reviewed October 29, 2026
Additional source	Hugging Face model metadata (retrieved October 29, 2026)
Additional source	Liquid AI LFM3-1.7B edge-inference chart (retrieved October 29, 2026)
Additional source	Model metadata (retrieved October 29, 2026)1 2 3 4