ProductTensorFlow Lite / TFLite / LiteRT

LiteRT LM

An on-device large model inference runtime introduced by Google, optimized for mobile CPUs, GPUs, and NPUs, supporting weight quantization, streaming output, and KV Cache reuse.

Timeline (last 90 days)

Oct 13

Anvil's chat feature uses the Gemma 4 E2B model, executed through the LiteRT LM runtime.

Unverified50%

Oct 13

LiteRT LM supports weight quantization (int8, int4) to compress model size, and provides inference acceleration mechanisms such as streaming output and KV Cache reuse.

Unverified50%

All Facts (2)

Unverified

Anvil's chat feature uses the Gemma 4 E2B model, executed through the LiteRT LM runtime.

50%Unverified

LiteRT LM supports weight quantization (int8, int4) to compress model size, and provides inference acceleration mechanisms such as streaming output and KV Cache reuse.

50%

Source Articles

Dual On-Device Models on iPhone: How a Chat Model and an Image Model CoexistOct 13