Models & Labs
vLLM v0.34.0rc0 adds Gemma and StableLM support
vLLM Releaseshigh confidence
Why it matters
- →Expands vLLM's supported model library to include Gemma and StableLM.
- →Enables high-throughput inference for these models using vLLM's optimized backend.
- →Reduces friction for teams adopting open-weight models from Google and Stability AI.
vLLM has released version 0.34.0rc0, a release candidate that extends its Transformers backend to support Google's Gemma and Stability AI's StableLM models. The update, authored by Harry Mellor with co-authorship from Claude Opus 5.5, integrates these architectures into vLLM's high-throughput serving engine. This expansion allows users to deploy Gemma and StableLM using vLLM's existing optimization features, such as PagedAttention and continuous batching. The release underscores vLLM's strategy of broadening support for open-weight models to maintain its position as the standard inference runtime.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1llama.cpp b11185 adds ROCm 10 and CUDA 13
llama.cpp Releases · September 26, 2026 · Same story
- 2llama.cpp b11606 release adds ROCm 10 and CUDA 13
llama.cpp Releases · October 14, 2026 · Same story
More in Models & Labs
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
llama.cpp Releases
Models & Labscoding
llama.cpp b12342 adds few-row MMQ for speculative decoding
Speculative decoding just got a significant speed boost on NVIDIA GPUs. This release introduces a specialized kernel for 'few-row' matrix multiplications, targeting the small batch sizes typical of verification passes in speculative decoding workflows. By optimizing these specific operations for quantized formats like Q8_0 and IQ4 variants, llama.cpp reduces latency where it matters most during token generation. The change is CUDA-only for now but leverages hardware capabilities from Ampere onwards, making local inference faster without requiring new model architectures or complex configuration changes.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google deploys Gemini 4 Hydrogen for faster Search
Google is quietly shifting its search infrastructure by replacing older fast models with Gemini 4 Hydrogen, a tier designed specifically for speed and cost-efficiency rather than raw reasoning power. This model completes tasks 2.5 times faster than its predecessor while beating the more powerful Argon on specific intelligence benchmarks, proving that specialized efficiency can outperform brute force in high-volume workflows. By integrating this into Search's fan-out technique, Google now breaks down complex queries into subqueries with greater precision and speed, delivering richer visual responses without the latency of heavier models. This marks a strategic pivot toward agentic execution layers within consumer search, making AI Mode feel snappier for everyday multi-part questions.
TechCrunch AI