Top story
runtime release ·
MLX-LM v0.32.0
Honour the left-padding mask in MiniCPM3 and Phixtral batches by @bhaskargurram-ai in https://github.com/ml-explore/mlx-lm/pull/1925
Open models · sources checked 16 Oct 2026
AI you can download, run and keep.
Open models are published as files anyone can download: you run them on your own hardware instead of sending your work to someone else’s. Mutinai tracks the models, the machines they fit and the tools that run them, and labels every figure with where it came from.
- Something like ChatGPT, on my computer?
- What’s good for coding?
- Can a Mac run these?
- Which GPU should I buy?
- What does quantization mean?
- 25open-weight models
- 12releases & launches in 30 days
- Best open GPQA Diamond 71.5%(DeepSeek-R1)
- 14tools & runtimes
What’s happening right now
Latest
- Toolruntime release
- EventannouncementScaling Indic Pretraining: 2.8 Trillion Tokens and an Open Data Generation Pipeline
- EventannouncementGranite Switch 4.1: Chat, Decisions, and Speech from One Endpoint
- Toolruntime release
- EventannouncementMistral AI
- EventannouncementHiRO-ACEv2: Simulating extreme weather at 3 km resolution for a fraction of the cost
- EventannouncementRun fast, private agents on-device with LFM2.5-24B-A2B
- Toolruntime release
Models worth knowing
Open models arrive in families. Start with the family and who builds it, then open a release to see its sizes and variants.
Qwen
Qwen Team (Alibaba Cloud)Known forCoding, reasoning and tool use596M – 235B parametersHow many numbers the model learned during training — the usual rough measure of its size.Dense and mixture of expertsA model split into parts where only a few run for each word, so it answers faster than its size suggests.Newest release 5 Aug 2026Gemma
Google DeepMindKnown forImages, long documents and many languages9.24B – 27.4B parametersHow many numbers the model learned during training — the usual rough measure of its size.DenseNewest release 12 Mar 2025Mistral
Mistral AIKnown forTool use and many languages7.25B – 23.6B parametersHow many numbers the model learned during training — the usual rough measure of its size.DenseNewest release 30 Jan 2025DeepSeek-R1
DeepSeekKnown forCoding, reasoning and long documents671B parametersHow many numbers the model learned during training — the usual rough measure of its size.Mixture of expertsA model split into parts where only a few run for each word, so it answers faster than its size suggests.Newest release 20 Jan 2025Phi
MicrosoftKnown forCoding and reasoning14.7B parametersHow many numbers the model learned during training — the usual rough measure of its size.DenseNewest release 12 Dec 2024Llama
MetaKnown forCoding, tool use and long documents8.03B – 70.6B parametersHow many numbers the model learned during training — the usual rough measure of its size.DenseNewest release 6 Dec 2024
How they compare
A benchmarkA fixed set of questions every model is given, so their scores can be compared on the same task. is a fixed set of questions every model answers, so scores can be compared. Positions below are relative to the best open result Mutinai holds for that test.
Coding best scores on HumanEval
- Qwen2.5-Coder 32B92.7
- Llama 3.3 70B88.4
- Qwen2.5 32B88.4
developer-reported what it measures →
Reasoning best scores on GPQA Diamond
- DeepSeek-R1 671B71.5
- Qwen3 30B-A3B65.8
- Qwen2.5 32B62.1
developer-reported what it measures →
Knowledge best scores on MMLU-Pro
- DeepSeek-R1 671B84.0
- Phi-4 14B70.4
- Qwen2.5 32B69.0
developer-reported what it measures →
Instruction following best scores on IFEval
- Llama 3.3 70B92.1
- Gemma 3 27B90.4
- Llama 3.1 8B80.4
developer-reported what it measures →
Runs on your own machine
- Qwen2.5-Coder 32B~21 GB to runruns well on 11 of 19 reference systems
- Qwen2.5-Math 7B~6 GB to runruns well on 17 of 19 reference systems
- Qwen3 30B-A3B~18 GB to runruns well on 11 of 19 reference systems
estimated smallest download at 8K contextHow much text the model can consider at once, counted in tokens — roughly ¾ of a word each.
Images and multimodal
1 tracked model accepts images, but Mutinai holds no multimodal benchmark results yet, so there is nothing to rank here. See the models →
Hardware watch
Memory decides what fits; memory speed decides how fast it answers. Prices are shown only with what they mean and when they were checked.
Graphics memoryThe memory on a graphics card. A model has to fit in it to run at full speed. tiers
- 80GBHolds models up to ~120B at 4-bit estimatedA100 80GB
- 32GBHolds models up to ~47B at 4-bit estimatedRTX 5090$1,999 MSRP
- 24GBHolds models up to ~35B at 4-bit estimatedRTX 4090$1,599 MSRPRX 7900 XTX$999 MSRPRTX 3090$1,499 MSRP
- 16GBHolds models up to ~22B at 4-bit estimatedRTX 4060 Ti 16GB$499 MSRP
Mutinai records launch prices (MSRPThe price the manufacturer set at launch. What a shop charges today can be very different.) and, where a retail price has been checked, the date it was checked. It keeps no price history yet, so it shows no price trends.
Unified memoryOne pool of memory shared by the processor and graphics, as on Apple silicon, so large models can fit.
These chips share one pool of memory with the processor, so what fits is chosen when the machine is bought, not by the chip.
What the community is talking about
Community Runs and reviews written by members will appear here, kept separate from the sourced facts everywhere else on this page and never merged into them.
Not collected yet
Contributions are not open yet
This preview has no sign-in, so nothing here could have been written by a member: the accounts, runs and reviews in its database are fictional sample content, and are left out rather than dressed up as activity. Real runs, reviews and discussion appear here once contributions open.
Why run models yourself
Your data stays put
Prompts, code and documents stay on the machine you ran them on.
How local models work →No account, no per-token bill
Download the weights once and run them on a laptop, one graphics card, a Mac or a server.
What can I run? →Nobody can take it back
The weights are files you keep. Nobody can withdraw, reprice or quietly change them.
What “open” means →