Daily AI news, sourcedUpdates / Models
Updates Models · Unconfirmed · 219 items · newest issue first · page 3 of 11
AreaAllModelsAgents & codingApps & productsChips & infraResearchPricing & limitsBusiness & fundingPolicy & safetyOther
- User reports running a 4-bit Qwen3.8-27B with 262k context and vision on a 16GB GPU at 80 tokens per secondUnconfirmedr/LocalLLaMAOct 13Models
- Precision Numerics releases Boole-1.7B, a 1-bit Qwen3-1.7B it says beats Bonsai-1.7BUnconfirmedr/LocalLLaMAOct 13Models
- vLLM fork serves Qwen3.8-Flash-Next to 7 users at 262k context on six RTX 3090sUnconfirmedr/LocalLLaMAOct 12Models
- Bit-exact speculative decoding hits 106 t/s on Qwen3.8-27B with one RTX PRO 6000Unconfirmedr/LocalLLaMAOct 12Models
- Unverified posts claim a Claude Next model is being tested inside Claude CodeUnconfirmedX @chetasluaOct 12Models
- Elon Musk says Grok 4.8 is a 2.5T model on a new C++ stack finishing training then starting RLUnconfirmedX @elonmuskOct 12Models
- LocalLLaMA user proposes a C# internet-radio player prompt to judge local model qualityUnconfirmedr/LocalLLaMAOct 12Models
- mlx-community publishes SenseNova-U1.5-8B-MoT 8-step V2 8-bit for Apple Silicon image generationUnconfirmedHugging Face @mlx-communityOct 12Models
- LocalLLaMA post ranks models that fit 10–16GB VRAM for more spontaneous local chatUnconfirmedr/LocalLLaMAOct 12Models
- mlx-community publishes SenseNova-U1.5-8B-MoT 8-step V2 4-bit for Apple Silicon image generationUnconfirmedHugging Face @mlx-communityOct 12Models
- LocalLLaMA user runs Qwen3.8-Flash-Next at 500k context on an AMD Radeon RX 7900 XTXUnconfirmedr/LocalLLaMAOct 12Models
- LocalLLaMA user shares vllm-ascend progress running GLM-5.3-Flash on dual Ascend 310P cardsUnconfirmedr/LocalLLaMAOct 12Models
- mlx-community posts a 6-bit Apple Silicon port of JetBrains Mellum2.1-12B-A2.5B-ThinkingUnconfirmedHugging Face @mlx-communityOct 11Models
- mlx-community publishes a Gemma 4 E2B OptiQ 4-bit decision model for Apple SiliconUnconfirmedHugging Face @mlx-communityOct 11Models
- llama.cpp PR #29416 adds MiniCPM-V 4.7 model supportUnconfirmedr/LocalLLaMAOct 11Models
- LocalLLaMA user says Qwen3.8-Flash-Next on Strata with pi beat Claude Code Opus 5.5 on an everyday-tasks bench on a 5090Unconfirmedr/LocalLLaMAOct 11Models
- LocalLLaMA user hits SWE-Bench LimitExceeded with Gemma 4’s agentic chat template vs Qwen3.8 27BUnconfirmedr/LocalLLaMAOct 11Models
- LocalLLaMA user runs Qwen3.8-Flash-Next-oQ4e-mtp on an M3 Max 64GB Mac with oMLXUnconfirmedr/LocalLLaMAOct 11Models
- User reports ~80 tok/s for Qwen3.6-35B-A3B on a 64GB Strix Halo mini PCUnconfirmedr/LocalLLaMAOct 11Models
- LocalLLaMA user benches three Qwen 3.8 quants on recent AtCoder problems across four thinking effortsUnconfirmedr/LocalLLaMAOct 11Models