Liquid AI model product
LFM2.5 2.6B
6B dense hybrid text→text model (LFM2 architecture) post-trained for on-device agentic workloads: planning, native tool calling, and multi-step tasks.
Updated Aug 17, 2026. Default version: LFM2.5-2.6B
LLMBoard Score32.6LFM2.5-2.6B
Coverage60%8 benchmark families
Context windowN/ATokens
Official input priceN/AOfficial price unavailable
LFM2.5 2.6B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
LFM2.5 2.6B Benchmark Results
Benchmark scores for LFM2.5-2.6B.
Sort by
Benchmark | Score | Rank | Participants | Percentile | Evidence | Evaluated |
|---|
| BenchmarkAA-Omniscience Index | Score-29.5 points | Rank02 | Participants2 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkMulti-IF | Score80.1% | Rank02 | Participants23 | Percentile95.5% | EvidenceC | Evaluated |
| BenchmarkPinchBench | Score68.2% | Rank06 | Participants6 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkTau3 Banking | Score5.7% | Rank07 | Participants7 | Percentile0.0% | EvidenceC | Evaluated |
| BenchmarkClaw-Eval | Score62.8% | Rank08 | Participants14 | Percentile46.1% | EvidenceC | Evaluated |
| BenchmarkBFCL-V4 | Score56.9% | Rank11 | Participants15 | Percentile28.6% | EvidenceC | Evaluated |
| BenchmarkIFBench | Score59.2% | Rank29 | Participants34 | Percentile15.2% | EvidenceC | Evaluated |
| BenchmarkLiveCodeBench v6 | Score59.4% | Rank40 | Participants56 | Percentile29.1% | EvidenceC | Evaluated |
| BenchmarkAIME 2025 | Score51.9% | Rank100 | Participants115 | Percentile13.2% | EvidenceC | Evaluated |
LFM2.5 2.6B Arena Results
Preference and agent-evaluation results for the default version.
No Arena results
The default version does not have a matching Arena result yet.
LFM2.5 2.6B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.10 input, $0.20 output per 1M via NanoGPT
- Tracked offerings
- 1
Sort by
Provider | Provider model ID | Region | Input / 1M | Output / 1M | Context | Updated |
|---|
| ProviderNanoGPT | Provider model IDliquid/lfm-2.5-2.6b | Regionglobal | Input / 1M$0.10 | Output / 1M$0.20 | Context128K | Updated |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
LFM2.5 2.6B Runtime Performance
Provider-specific output speed and catalog latency for LFM2.5-2.6B. Runtime does not affect the capability score.
No runtime data
Browse runtime rankingsNo provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
LFM2.5 2.6B Specifications
Technical details for the model's default version.
- Version
- LFM2.5-2.6B
- Released
- Aug 4, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 2.7B
- Context window
- N/A
- Max output
- N/A
- Inputs
- text
- Outputs
- text
- Open weights
- No
- License
- LFM Open License v1.0
LFM2.5 2.6B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
Sort by
Version | Released | LLMBoard | Parameters | Context | Max output | Open weights | License |
|---|
| VersionLFM2.5-2.6B | Released | LLMBoard32.6 | Parameters2.7B | ContextN/A | Max outputN/A | Open weightsNo | LicenseLFM Open License v1.0 |
LFM2.5 2.6B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to LFM2.5 2.6B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
LA
LFM2.5 VL 3B
Liquid AI
MA
Mistral Small 4
Mistral AI
ME
LongCat Flash Lite
Meituan
AC
Qwen3.5 4B
Alibaba Cloud / Qwen Team
SA
Sarvam 30B
Sarvam AI
GO
Google
DiffusionGemma 26B A4B
What is LFM2.5 2.6B?
Key information about LFM2.5 2.6B and its available data.
6B dense hybrid text→text model (LFM2 architecture) post-trained for on-device agentic workloads: planning, native tool calling, and multi-step tasks. 69B parameters, trained on ~34T tokens, 128K context (131,072 tokens), 128K vocabulary. Always thinks before answering. cpp, MLX, and ONNX. 6B (+ Base). 0). 50 (signed).
49 — recorded here only. 6B. 6b:free exists externally but was not added.
Data as of 2026-08-17.
FAQ
Common questions about LFM2.5 2.6B.
When was LFM2.5 2.6B released?
LFM2.5 2.6B's default version was released on Aug 4, 2026.
How much does LFM2.5 2.6B cost?
No official standard PAYG price is currently available for LFM2.5 2.6B. The lowest tracked third-party offer starts at $0.10 input and $0.20 output via NanoGPT.
Who created LFM2.5 2.6B?
LFM2.5 2.6B was created by Liquid AI.
What is the context window for LFM2.5 2.6B?
A context window is not available for the default version.
Is LFM2.5 2.6B open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer LFM2.5 2.6B?
1 provider offerings are linked to the default version.
What models should I compare LFM2.5 2.6B with?
Nearby ranked alternatives include o1 pro, DiffusionGemma 26B A4B, Sarvam 30B.