whatmodel.sh · find the best AI models you can run

Best AI models for web search to run locally on a 32GB Mac

A leaderboard of the best AI models for web search you can run locally on a Mac with 32GB of memory. Mistral Small 3.1 24B (2503) is #1 right now, followed by Muse Glimmer 30B and Qwen3.6 27B. Models are ranked on web research tests like BrowseComp, with search turned on. Each model shows the download that fits and the command to run it.

23 models · scores combine 12 benchmarks · updated how it works · find a leaderboard · benchmark your machine
task[all tasks][agent tasks][coding][chat & writing][math][vision & documents][web search]

Scores use reviewed web search evaluations only. Compare scores within this task. Models without task measurements are excluded.

Models for 32GB RAM · Apple Silicon → ~21GB fast · local only · reset
+ more
chip[M1 Pro][M1 Max][M2 Pro][M2 Max][M4][M5][M6 (24/32GB)]
set[all sets][Arena search systems (style controlled) · 96][BrowseComp-ZH · 13][BrowseComp-ZH · 5][BrowseComp-ZH · 3][Humanity's Last Exam with tools · 3][BrowseComp with context management · 3][BrowseComp with context management · 2][Humanity's Last Exam with tools · 2][Humanity's Last Exam with tools · 2][BrowseComp · 2]

Filter the model list while keeping task scores. Models without a shared score scale are not yet calibrated; select a set to see its own scores.

Detect your exact chip or graphics card:

curl -G whatmodel.sh/ -d max=0 --data-urlencode "mac=$(sysctl -n hw.memsize machdep.cpu.brand_string)"Mac

curl -G whatmodel.sh/ -d max=0 --data-urlencode "ram=$(awk '/MemT/{print $2}' /proc/meminfo)" --data-urlencode "nv=$(nvidia-smi --query-gpu=name,memory.total --format=csv,noheader 2>/dev/null)"Linux

curl.exe -G whatmodel.sh/ -d max=0 --data-urlencode "ram=$((gcim Win32_ComputerSystem).TotalPhysicalMemory)" --data-urlencode "nv=$(nvidia-smi --query-gpu=name,memory.total --format=csv,noheader 2>$null)"Windows

$ terminal

detecting your OS…

Run curl "whatmodel.sh/?ram=32&gpu=apple&max=0"to get this list in your terminal (add &format=json for agents).

ram[8][16][24][32][48][64][96][128][192][256]GB
gpu[apple silicon][graphics card][none]
price[local][$0.05][$0.25][$1][$5][$20][any]max per M

detecting your OS…

Score vs memory to run

#modeltask score↓where↓speed↓RAM↓context↓get itsources
#modeltask score↓where↓speed↓RAM↓context↓get itsources
16Mistral Small 3.1 24B (2503) · 4-bit Mistral62±3*local~17GB128kLM Studio › Finance Agent Benchmark (Vals AI, arXiv:2508.00828v1)
52Muse Glimmer 30B · 4-bit Meta48±6*†local~21GB131kOllama › meta-models/Muse-Glimmer-30B model card (speed and benchmark tables)
56Qwen3.6 27B · 4-bit Alibaba46±6*local~20GB262kOllama › meta-models/Muse-Glimmer-30B model card (speed and benchmark tables)
71Qwen3.5 27B · 4-bit Alibaba41±3†local~20GB262kOllama › Qwen/Qwen3.5-122B-A10B model cardQwen/Qwen3.6-35B-A3B model card
76Gemma 4 31B · 4-bit Google38±3local~21GB262kOllama › K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts (arXiv:2606.02404v1)meta-models/Muse-Glimmer-30B model card (speed and benchmark tables)google/gemma-4-31B-it model card
88Gemma 4 26B-A4B · 4-bit Google23±8*†local~19GB262kllama.cpp › google/gemma-4-31B-it model card
—Gemma 4 12B · Q8_0 Googlenot yet calibratedlocal~15GB262kllama.cpp › XHToken/Spark-X2.5-4B model card
—Gemma 4 E2B · mlx-8bit Googlenot yet calibratedlocal~7.7GB131kLM Studio › openbmb/MiniCPM5-2B model cardXHToken/Spark-X2.5-4B model card
—Gemma 4 E4B · mlx-8bit Googlenot yet calibratedlocal~11GB131kLM Studio › openbmb/MiniCPM5-2B model cardXHToken/Spark-X2.5-4B model card
—GLM-4.7-Flash · Q4_K_M Z.ainot scored yetlocal~21GB200kllama.cpp ›
—gpt-oss-20b · mxfp4 OpenAInot yet calibratedlocal~15GB131kOllama › OpenAI — gpt-oss-120b & gpt-oss-20b Model Card
—Granite 4.2 3B · Q8_0 IBMnot yet calibratedlocal~5.6GB131kllama.cpp › openbmb/MiniCPM5-2B model card
—K2 Horizon 7B · gguf-q8_0 Institute of Foundation Models (MBZUAI)not scored yetlocal~13GB524kllama.cpp ›
—LFM2.5 2.6B · mlx-8bit Liquid AInot yet calibratedlocal~4.5GB131kLM Studio › openbmb/MiniCPM5-2B model card
—LFM2.5 8B-A1B · q8_0 Liquid AInot yet calibratedlocal~13GB128kOllama › openbmb/MiniCPM5-2B model card
—MiniCPM5 2B · mlx-8bit OpenBMBnot yet calibratedlocal~4.3GB131kLM Studio › openbmb/MiniCPM5-2B model card
—Nemotron 3 Nano 30B-A3B · mlx-4bit NVIDIAnot scored yetlocal~20GB262kLM Studio ›
—Nemotron 3 Nano 4B · q8_0 NVIDIAnot yet calibratedlocal~7GB262kOllama › openbmb/MiniCPM5-2B model card
—Qwen3.5 2B · q8_0 Alibabanot yet calibratedlocal~4.9GB262kOllama › openbmb/MiniCPM5-2B model cardXHToken/Spark-X2.5-4B model card
—Qwen3.5 4B · mlx-8bit Alibabanot yet calibratedlocal~6.9GB262kLM Studio › openbmb/MiniCPM5-2B model cardXHToken/Spark-X2.5-4B model card
—Qwen3.5 9B · mlx-8bit Alibabanot yet calibratedlocal~12GB262kLM Studio › XHToken/Spark-X2.5-4B model card
—Spark-X2.5 1.7B · bf16 iFLYTEK (SparkLLM)not yet calibratedlocal~5.1GB1MOllama › XHToken/Spark-X2.5-4B model card
—Spark-X2.5 4B · bf16 iFLYTEK (SparkLLM)not yet calibratedlocal~10GB1MOllama › XHToken/Spark-X2.5-4B model card

+ 108 more don't fit your machine or budget · show them

131 of 135 catalogue models listed here · sources and licences

* new or thinly tested: ranked on too few benchmark results · † only the maker’s own numbers so far · ~RAM is estimated; full context may need more memory