Timeline of Liquid AI

LiquidAI/LFM2.5-VL-3B-GGUF: parameter count changed from 3000000000 to 2697198592

LiquidAI/LFM2.5-VL-1.6B-GGUF: parameter count changed from 1600000000 to 1170340608

LiquidAI/LFM2.5-8B-A1B-DSpark-GGUF: parameter count changed from 8000000000 to 327707521

LiquidAI/LFM2.5-350M-GGUF: parameter count changed from 350000000 to 354483968

LiquidAI/LFM2.5-2.6B-DSpark-GGUF: parameter count changed from 2600000000 to 327707521

LiquidAI/LFM2.5-1.2B-Instruct-GGUF: parameter count changed from 1200000000 to 1170340608

LiquidAI/LFM2.5-8B-A1B-GGUF: parameter count changed from 8000000000 to 8467856832

LiquidAI/LFM2.5-230M-GGUF: parameter count changed from 230000000 to 229693184

LiquidAI/LFM2.5-2.6B-GGUF: parameter count changed from 2600000000 to 2697198592

New model: lfm-40b (Liquid AI)

New model: lfm2-5-1-2b-thinking (Liquid AI)

New model: lfm2-8b-a1b (Liquid AI)

New model: lfm2-2-6b (Liquid AI)

New model: lfm2-24b-a2b (Liquid AI)

LFM2-1.2B scores 0% on Terminal-Bench

LFM2-1.2B scores 12.57% on τ²-bench

LFM2-1.2B scores 21.97% on IFBench

LFM2-1.2B scores 5.56% on Humanity's Last Exam

LFM2-1.2B scores 22.83% on GPQA

LFM2-1.2B scores 4.83 on Artificial Analysis Intelligence Index

lfm-40b scores 4.93% on Humanity's Last Exam

lfm-40b scores 32.73% on GPQA

lfm-40b scores 5.42 on Artificial Analysis Intelligence Index

LFM2.5-VL-1.6B scores 0% on Terminal-Bench

LFM2.5-VL-1.6B scores 8.48% on τ²-bench

LFM2.5-VL-1.6B scores 26.53% on MMMU-Pro

LFM2.5-VL-1.6B scores 33.13% on IFBench

LFM2.5-VL-1.6B scores 5.1% on Humanity's Last Exam

LFM2.5-VL-1.6B scores 28.89% on GPQA

LFM2.5-VL-1.6B scores 4.83 on Artificial Analysis Intelligence Index

LFM2.5-1.2B-Instruct scores 0% on Terminal-Bench

LFM2.5-1.2B-Instruct scores 10.82% on τ²-bench

LFM2.5-1.2B-Instruct scores 43.81% on IFBench

LFM2.5-1.2B-Instruct scores 6.72% on Humanity's Last Exam

LFM2.5-1.2B-Instruct scores 32.63% on GPQA

LFM2.5-1.2B-Instruct scores 5.21 on Artificial Analysis Intelligence Index

lfm2-5-1-2b-thinking scores 0% on Terminal-Bench

lfm2-5-1-2b-thinking scores 19.59% on τ²-bench

lfm2-5-1-2b-thinking scores 41.84% on IFBench

lfm2-5-1-2b-thinking scores 6.21% on Humanity's Last Exam

lfm2-5-1-2b-thinking scores 33.94% on GPQA

lfm2-5-1-2b-thinking scores 5.22 on Artificial Analysis Intelligence Index

lfm2-8b-a1b scores 0% on Terminal-Bench

lfm2-8b-a1b scores 10.53% on τ²-bench

lfm2-8b-a1b scores 26.26% on IFBench

lfm2-8b-a1b scores 4.91% on Humanity's Last Exam

lfm2-8b-a1b scores 34.44% on GPQA

lfm2-8b-a1b scores 4.93 on Artificial Analysis Intelligence Index

lfm2-2-6b scores 0.76% on Terminal-Bench

lfm2-2-6b scores 13.45% on τ²-bench

lfm2-2-6b scores 19.52% on IFBench

lfm2-2-6b scores 5.47% on Humanity's Last Exam

lfm2-2-6b scores 30.61% on GPQA

lfm2-2-6b scores 5.21 on Artificial Analysis Intelligence Index

LFM2.5-2.6B (free) scores 4.49% on Terminal-Bench

LFM2.5-2.6B (free) scores 14.35% on SciCode

LFM2.5-2.6B (free) scores 6.21% on Humanity's Last Exam

LFM2.5-2.6B (free) scores 55.76% on GPQA

LFM2.5-2.6B (free) scores 8.39 on Artificial Analysis Intelligence Index

LFM2.5-8B-A1B scores 4.55% on Terminal-Bench

LFM2.5-8B-A1B scores 16.08% on τ²-bench

LFM2.5-8B-A1B scores 55.65% on IFBench

LFM2.5-8B-A1B scores 6.86% on Humanity's Last Exam

LFM2.5-8B-A1B scores 51.31% on GPQA

LFM2.5-8B-A1B scores 7.22 on Artificial Analysis Intelligence Index

lfm2-24b-a2b scores 0% on Terminal-Bench

lfm2-24b-a2b scores 11.11% on τ²-bench

lfm2-24b-a2b scores 45.85% on IFBench

lfm2-24b-a2b scores 4.22% on Humanity's Last Exam

lfm2-24b-a2b scores 47.37% on GPQA

lfm2-24b-a2b scores 5.95 on Artificial Analysis Intelligence Index

New model: LFM2.5-Audio-1.5B (Liquid AI)

LiquidAI/LFM2.5-Audio-1.5B-GGUF: parameter count changed from 1170340608 to 1500000000

LiquidAI/LFM2.5-VL-3B-GGUF: parameter count changed from 2697198592 to 3000000000

LiquidAI/LFM2.5-VL-1.6B-GGUF: parameter count changed from 1170340608 to 1600000000

New model: LFM2.5-230M-Base (Liquid AI)

New model: LFM2.5-8B-A1B-DSpark (Liquid AI)

LiquidAI/LFM2.5-8B-A1B-DSpark-GGUF: parameter count changed from 327707521 to 8000000000

New model: LFM2.5-8B-A1B-Base (Liquid AI)

LiquidAI/LFM2.5-350M-GGUF: parameter count changed from 354483968 to 350000000

New model: LFM2.5-350M-Base (Liquid AI)

New model: LFM2.5-2.6B-Base (Liquid AI)

New model: LFM2.5-2.6B-DSpark (Liquid AI)

LiquidAI/LFM2.5-2.6B-DSpark-GGUF: parameter count changed from 327707521 to 2600000000

New model: LFM2.5-1.2B-Base (Liquid AI)

LiquidAI/LFM2.5-1.2B-Instruct-GGUF: parameter count changed from 1170340608 to 1200000000

LiquidAI/LFM2.5-8B-A1B-GGUF: parameter count changed from 8467856832 to 8000000000

LiquidAI/LFM2.5-230M-GGUF: parameter count changed from 229693184 to 230000000

LiquidAI/LFM2.5-2.6B-GGUF: parameter count changed from 2697198592 to 2600000000

New model: LFM2.5-VL-3B (Liquid AI)

New model: LiquidAI/LFM2.5-Audio-1.5B-GGUF (Liquid AI)

New model: LiquidAI/LFM2.5-VL-3B-GGUF (Liquid AI)

New model: LFM2-VL-1.6B (Liquid AI)

New model: LFM2.5-VL-450M (Liquid AI)

New model: LiquidAI/LFM2.5-VL-1.6B-GGUF (Liquid AI)

New model: LFM2.5-230M (Liquid AI)

New model: LiquidAI/LFM2.5-8B-A1B-DSpark-GGUF (Liquid AI)

New model: LFM2-1.2B (Liquid AI)

New model: LFM2.5-8B-A1B (Liquid AI)

New model: LiquidAI/LFM2.5-350M-GGUF (Liquid AI)

New model: LFM2.5-350M (Liquid AI)

New model: LiquidAI/LFM2.5-2.6B-DSpark-GGUF (Liquid AI)

New model: LFM2.5-VL-1.6B (Liquid AI)

New model: LFM2.5-1.2B-Instruct (Liquid AI)

New model: LiquidAI/LFM2.5-1.2B-Instruct-GGUF (Liquid AI)

New model: LiquidAI/LFM2.5-8B-A1B-GGUF (Liquid AI)

New model: LiquidAI/LFM2.5-230M-GGUF (Liquid AI)

New model: LiquidAI/LFM2.5-2.6B-GGUF (Liquid AI)

New model: LFM2.5-2.6B (free) (Liquid AI)

OpenRouter lists LFM2.5-2.6B (free) at $0 in / $0 out per 1M tokens

New company: Liquid AI