Microsoft
Maker · 21 entries
Everything from Microsoft
| Name | What to know | Price or memory | |
|---|---|---|---|
| mixed Strong at image editing and cheap, but trails GPT Image Sunburst on Arena | |||
| New open model on Hugging Face (text-generation): 2.29B parameters, 4.58 GB of weights, license mit, 4 community quantizations incl. GGUF | ~1.4 GB | ||
| Multi-model harness reports 71.2% on Terminal-Bench 4.1 at 88% lower cost than Claude Fable 5.2 | |||
| unproven Microsoft says it nearly matches top coding models cheaply; no independent tests yet | |||
| unproven Microsoft says it matches cheaper rivals on coding, but no one else has tested it | |||
| MAI Cyber 2 Flash for agentic codebase defense; output tokens 86% cheaper than Cyber 1.1 Flash; existing MDASH customers can move to it | |||
| Decision 1.1 routing model is generally available, used by 4,000 customers | |||
| unproven Cheap, fast image generation at about 1 cent each; quality not yet independently tested | |||
| unproven Announced Azure-only customization of MAI models; not downloadable, no price or date | |||
| MAI Realtime 1, Microsoft's first live speech-to-speech model, starts at $0.01 per audio minute; try it in MAI Labs | |||
| unproven Microsoft says it is the fastest and most accurate; no independent tests yet | |||
| Supports over 50 languages natively; finetuning code and vLLM inference are available | ~1.6 GB | ||
| promising Small, MIT-licensed speech model from Microsoft, but meant for research only | ~1.6 GB | ||
| Microsoft's Quicksand library bundles QEMU and Alpine Linux in a Python package; used by external-sandbox-quicksand | |||
| strong Established browser automation framework that newer tools are compared against and still preferred | |||
| A small weight change can leave an agent normal in routine tests but altered under specific conditions; advice: pin model versions, verify signatures and hashes | |||
| Used as depth baseline that Marigold v2 outperforms | |||
| Unified streaming ASR continuously transcribes who said what as speech arrives, with custom hotwords and 10 languages | |||
| Fara1.5 released in 4B, 9B and 27B sizes, a much improved successor to the earlier Fara-7B web computer-use agent | ~76 GB | ||
| Compact 4B-parameter model for text-to-image generation and instruction-based image editing, runnable locally on one GPU | ~17.5 GB | ||
| Added experimental speakers: multilingual voices in nine languages and 11 English style voices | |||