DeepSeek-V4-Flash vs LFM2.5-1.2B-Thinking
DeepSeek vs Liquid AI — benchmarks, pricing, and capabilities side by side.
- •DeepSeek-V4-Flash has the higher intelligence index (89.4 vs 33.9)
- •LFM2.5-1.2B-Thinking is cheaper ($0.00 vs $0.10 per 1M input)
| DeepSeek-V4-Flash | LFM2.5-1.2B-Thinking | |
|---|---|---|
| Intelligence index | 89.4 | 33.9 |
| Developer | DeepSeek | Liquid AI |
| Type | LLM | LLM |
| Access | Open weights | — |
| Context window | 1,048,576 tokens | — |
| Input price | $0.10 / 1M | $0.00 / 1M |
| Output price | $0.20 / 1M | $0.00 / 1M |
| Speed | 109 tok/s | — |
| Released | April 24, 2026 | January 20, 2026 |
| Parameters | 284B (13B active) | — |
| Input modalities | Text | — |
| Output modalities | Text | — |
Shared benchmarks
DeepSeek-V4-Flash
LFM2.5-1.2B-Thinking
GPQA Diamond
89.4
33.9
Humanity’s Last Exam
32.1
6.1