@0xSero: Best models smallest to largest right now. - Gemma-4-12B - Qwen3.8-27B - Laguna-S2.1 - Deepseek-V4-Flash - Inkling-Smal…
Summary
A tweet ranking the best open-weight AI models from smallest to largest, highlighting the vibrant open-weights community.
View Cached Full Text
Cached at: 08/03/26, 09:51 PM
Best models smallest to largest right now.
- Gemma-4-12B
- Qwen3.8-27B
- Laguna-S2.1
- Deepseek-V4-Flash
- Inkling-Small
- MiniMax-M3
- GLM-5.*
- Kimi-K2.7-Code
- Qwen3.8-Max
- Kimi-K3
Spoiled for choice, open weights community is much more exciting than the frontier. https://t.co/UcwSagLCev
Similar Articles
The new tiny open weight model Inkling 2 Mini nearly matches the Gemini 2.5 Pro
The tiny open-weight model Inkling 2 Mini, runnable locally with ~16-20 GB RAM, reportedly outperforms the cloud flagship Gemini 2.5 Pro on several benchmarks including SWE-bench Verified and ARC-AGI, showcasing rapid progress in compact model capabilities.
@jtdavies: Coding on small models... My default model for my 4xDGX Spark cluster is @UnslothAI's Qwen3.6-35B-A3B-NVFP4. I get exce…
A user tests various small AI models for coding tasks, finding Qwen3.6-27B-NVFP4 to be the best balance of speed and accuracy, and notes poor Java performance in these models.
XHToken/Spark-X2.5-4B VS inclusionAI/Ling-3.0-tiny VS Nanbeige/Nanbeige4.2-3B
The article asks users which small AI models are most useful and mentions three competing models in the same size class.
@nathanhabib1011: best models < 128B params on SWE-bench_pro... @Alibaba_Qwen 3.6 27b still crazy, closely followed by @ornith_ 35B
Tweet highlighting top-performing AI models under 128B parameters on the SWE-bench_pro benchmark, noting Alibaba Qwen 3.6 27B and ornith 35B as leading contenders.
Inkling by Thinking Machines is the #1 US open weight model now
Thinking Machines' Inkling model has achieved the top ranking among US open weight models.