Liquid AI · language model
LFM2.5-2.6B: pricing, context and benchmarks
LFM2.5-2.6B is a language model from Liquid AI released on August 11, 2026. It is offered free of charge on OpenRouter, with a 65.5K-token context window. No independent score has been published for it so far.
| Price | 30/30 | |
|---|---|---|
| Context | 2/15 | |
| Features | 15/15 |
What you need to know about LFM2.5-2.6B
LFM2.5-2.6B was released by Liquid AI (United States) on August 11, 2026. It accepts text as input and produces text. Its context window reaches 65,536 tokens, larger than 12 % of the models in the registry, and it can write up to 8,192 tokens in one go.
It is a recent model: real-world feedback and independent scores are still scarce.
LFM2.5-2.6B pricing
LFM2.5-2.6B is free on OpenRouter. Free variants usually have strict rate limits and can be withdrawn without notice: keep a paid fallback model for regular use.
Performance and benchmarks
Epoch AI does not yet publish a capabilities index for LFM2.5-2.6B. We therefore show no made-up performance rating: only price, context and features count towards its score.
Features, privacy and local use
- It supports tool calling, so it can drive an AI agent, call APIs or MCP servers.
- It can answer in an enforced JSON format (structured outputs).
- It is a reasoning model: it "thinks" before answering, which helps on hard tasks but uses more output tokens.
Its weights are open and downloadable from Hugging Face (other licence): you can run it on your own machines without sending your data to a third party. With 2,697,198,592 billion parameters, it needs about 1,507,177,162 GB of memory at 4-bit quantisation. Check with the VRAM calculator. It was downloaded 99,777 times on Hugging Face last month.
Other Liquid AI models
| Model | Released | Avg. price | ECI |
|---|---|---|---|
| LFM2.5-24B-A2B | Oct 13, 2026 | $0.21 | — |
Verdict
LFM2.5-2.6B scores 78/100 on our grid, which only uses published figures.
Strengths
- Usable for free on OpenRouter.
- Open weights: self-hostable, your data stays with you.
- Agent-ready: tool calling and structured JSON output.
Limitations
- No independent score published: performance cannot be verified.
- Text only: reads neither images nor files.
- Free version is rate-limited with no availability guarantee.
Avoid if…
- you want to host it on a single consumer computer (≈ 1,507,177,162 GB needed at 4-bit).
Alternatives
- Mistral Small 3$0.06
- Granite 4.0 Micro$0.04
- Llama 3.2 1B Instruct$0.07
- Llama 3 8B Lunaris$0.04
- Schematron V2 Turbo$0.06
- Command R7B (12-2024)$0.07
Comparisons and tools
Learn more
Frequently asked questions
How much does LFM2.5-2.6B cost?
LFM2.5-2.6B is free on OpenRouter as of October 25, 2026, with rate limits.
How big is LFM2.5-2.6B's context window?
65,536 tokens, roughly 49,000 words, with a maximum of 8,192 output tokens.
Can LFM2.5-2.6B run locally?
Yes, its weights are published under the other licence. Expect about 1,507,177,162 GB of memory at 4-bit.
Is LFM2.5-2.6B suitable for an AI agent?
Technically yes: it supports tool calling and structured outputs. Then check its cost for your volume.
What are the alternatives to LFM2.5-2.6B?
The closest models on price and performance are Mistral Small 3, Granite 4.0 Micro, and Llama 3.2 1B Instruct.
Related articles
- Mistral AI: models, pricing and agentsOctober 25, 2026
- LLM benchmarks: how to read model rankingsOctober 24, 2026
- Running a local LLM: hardware, software, and limitationsOctober 24, 2026
- Open-source or closed models for your AI agents?October 5, 2026
Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score