Is Liquid better than Recursive?
It depends on the job: Liquid offers the broader line-up (7 models vs 1), Recursive the longest context (262K tokens). The workload table above shows which wins for chat, RAG, agents, batch jobs and long documents.
Liquid vs Recursive
Liquid lists 7 models and Recursive 1. Below: line-up, price tiers, the cheapest model for each capability, and what each costs for five real workloads.
At a glance
| Liquid | Recursive | |
|---|---|---|
| Models listed | 7 | 1 |
| Models with published pricing | 7 | 1 |
| Typical price (median, 3:1 blend per 1M) | $0.03 | n/a |
| Cheapest paid model | LiquidAI: LFM2-2.6B ($0.01) | n/a |
| Longest context | LiquidAI: LFM2.5-24B-A2B (128K) | Recursive: Flint (free) (262K) |
| Free models | 3 | 1 |
| Newest model | LiquidAI: LFM2.5-24B-A2B (2026-10-13) | Recursive: Flint (free) (2026-10-21) |
| Price changes in the last 90 days | 0 | 0 |
Advertisement
What each costs for real workloads
For each workload: the cheapest model from each provider that can actually do it (fits the context window; tool calling required for the agent), and its monthly cost at list price.
| Workload (per month) | Liquid | Recursive |
|---|---|---|
| Customer-support chatbot 50,000 conversations — each a 6-turn conversation | $1.15 with LiquidAI: LFM2-2.6B | no suitable model |
| RAG search over documents 20,000 questions — each a question answered from ~8 retrieved passages | $1.40 with LiquidAI: LFM2-2.6B | no suitable model |
| Coding / tool-using agent 2,000 tasks — each a multi-step task with tool calls and file context | $14.40 with LiquidAI: LFM2.5-24B-A2B | no suitable model |
| Batch summarisation 10,000 documents — each a 6,000-word report summarised to a page | $0.92 with LiquidAI: LFM2-2.6B | no suitable model |
| Long-document analysis 500 documents — each a 150-page contract or codebase read in one pass | $3.20 with LiquidAI: LFM2.5-24B-A2B | no suitable model |
Cheapest model for each capability
| You need | Liquid | Recursive |
|---|---|---|
| Any model | LiquidAI: LFM2-2.6B ($0.01) | none listed |
| With tool calling | LiquidAI: LFM2.5-24B-A2B ($0.21) | none listed |
| With image input | none listed | none listed |
| With reasoning mode | LiquidAI: LFM2.5-24B-A2B ($0.21) | none listed |
| 128K+ context | LiquidAI: LFM2.5-24B-A2B ($0.21) | none listed |
| 1M+ context | none listed | none listed |
Line-up by type and price tier
| Model type | Liquid | Recursive |
|---|---|---|
| text->text | 7 | 1 |
| Price per 1M (3:1 blend) | Liquid | Recursive |
|---|---|---|
| Free | 3 | 1 |
| Under $0.50 | 4 | 0 |
| $0.50 – $3 | 0 | 0 |
| $3 – $10 | 0 | 0 |
| $10 and up | 0 | 0 |
Frequently asked questions
Which has more models, Liquid or Recursive?
Liquid: Liquid lists 7 models and Recursive lists 1.
Which provider has the longest context window?
Recursive: Liquid's longest is LiquidAI: LFM2.5-24B-A2B at 128K tokens; Recursive's is Recursive: Flint (free) at 262K.
Do Liquid or Recursive offer free models?
Liquid: 3 free models (e.g. LiquidAI: LFM2.5-1.2B-Instruct (free)). Recursive: 1 free model (e.g. Recursive: Flint (free)).
Which provider has models that can read images?
Liquid: 0 of 7 models accept images. Recursive: 0 of 1.