# LFM2.5 8B A1B vs Inkling FP4

LFM2.5 8B A1B is not currently in the Allocate serving catalog, so this page lists no prices for it: every price on this site comes from the live catalog.

## Specifications

| | LFM2.5 8B A1B | Inkling FP4 |
| --- | --- | --- |
| Lab | Liquid AI | Thinking Machines |
| Access | Not served on Allocate | Open weights |
| Context window | n/a | 512K tokens |
| List price, input | Not served | $1 / M tokens |
| List price, output | Not served | $4.05 / M tokens |
| Cached input | n/a | $0.17 / M tokens |
| License | Not listed | Apache 2.0 |
| Fine-tunable | Yes | Yes |

Specifications and provider list prices from the Allocate catalog, checked 2026-09-12.

## Choose LFM2.5 8B A1B for

- High-volume extraction
- Latency-sensitive routes
- Edge deployments

## Choose Inkling FP4 for

- Fine-tuning under a permissive license (Apache 2.0)
- Published cached-input pricing ($0.17 per M tokens)

## Common questions

### Which has the bigger context window?

Inkling FP4: 524,288 tokens (512K) against an unlisted window for LFM2.5 8B A1B.

### Can I fine-tune LFM2.5 8B A1B or Inkling FP4?

Both publish open weights (LFM2.5 8B A1B: Not listed; Inkling FP4: Apache 2.0), so both can be fine-tuned. On Allocate the trained weights stay inside your boundary and belong to you.

---

[HTML page](https://allocate.network/compare/liquid-lfm2-5-8b-a1b-vs-thinkingmachines-inkling) · [LFM2.5 8B A1B](https://allocate.network/models/liquid-lfm2-5-8b-a1b.md) · [Inkling FP4](https://allocate.network/models/thinkingmachines-inkling.md) · [Machine-readable catalog](https://allocate.network/catalog.json)