weckoai commited on
Commit
68fe44e
·
verified ·
1 Parent(s): cb015f7

Sync 2026-10-26: 172 models

Browse files
Files changed (3) hide show
  1. README.md +1 -1
  2. models.csv +3 -0
  3. models.json +60 -3
README.md CHANGED
@@ -36,7 +36,7 @@ table at **[modelfit.io/data](https://modelfit.io/data/)**; the canonical
36
  machine-readable source is
37
  **[modelfit.io/api/dataset](https://modelfit.io/api/dataset/)**.
38
 
39
- - **169 models** across 32 families (126 with a registry-verified local build, 26 open-weight models without a consumer-tier Ollama build, 17 closed cloud APIs tracked for comparison)
40
  - **License:** CC BY 4.0: reuse freely, including commercially, with attribution
41
  - **Updated:** see the `updated` field in `models.json` (regenerated from the live endpoint)
42
 
 
36
  machine-readable source is
37
  **[modelfit.io/api/dataset](https://modelfit.io/api/dataset/)**.
38
 
39
+ - **172 models** across 32 families (129 with a registry-verified local build, 26 open-weight models without a consumer-tier Ollama build, 17 closed cloud APIs tracked for comparison)
40
  - **License:** CC BY 4.0: reuse freely, including commercially, with attribution
41
  - **Updated:** see the `updated` field in `models.json` (regenerated from the live endpoint)
42
 
models.csv CHANGED
@@ -168,3 +168,6 @@ AREX-2 27B,AREX,27.4,Q4_K_M,32,17.2,64,true,true,false,ollama|llama.cpp|lm-studi
168
  Muse Spark Open,Muse,816,API,0,0,,false,true,false,,"Agentic coding, Multimodal, Long context",
169
  MiniMax M3.1 Flash,MiniMax,428,API,0,0,,false,true,false,,"Coding, Agentic, Computer use",
170
  Beam 1.0,Beam,501,API,0,0,,false,true,false,,"Agentic coding, Reasoning, Long context",
 
 
 
 
168
  Muse Spark Open,Muse,816,API,0,0,,false,true,false,,"Agentic coding, Multimodal, Long context",
169
  MiniMax M3.1 Flash,MiniMax,428,API,0,0,,false,true,false,,"Coding, Agentic, Computer use",
170
  Beam 1.0,Beam,501,API,0,0,,false,true,false,,"Agentic coding, Reasoning, Long context",
171
+ NVIDIA Nemotron 3.5 Bolt E4B,Nemotron,4,Q4_K_M,12,7.9,,true,true,false,ollama|llama.cpp|lm-studio,"Voice assistants, Tool calling, Chat",ollama run nemotron-3.5-bolt:e4b
172
+ Granite Switch 4.1 8B,Granite,8.8,Q4_K_M,12,6.4,,true,true,false,ollama|llama.cpp|lm-studio,"Chat, Decisions, Safety checks",
173
+ Granite Switch 4.1 30B,Granite,30,Q4_K_M,32,21,,true,true,false,ollama|llama.cpp|lm-studio,"Chat, Decisions, Safety checks",
models.json CHANGED
@@ -11,10 +11,10 @@
11
  "note": "A model fits a device when its estimatedLoadGb is within the memory budget: ~70% of unified memory up to 32GB, scaling linearly to ~85% at 128GB and above (high-RAM Macs can wire more memory to the GPU). Q4_K_M needs roughly 0.6 GB per billion parameters.",
12
  "fieldSemantics": "runsLocally = a registry-verified Ollama build fits at least one consumer RAM tier tracked here (up to 256GB). openWeights = the weights are publicly downloadable. Some open-weight giants (e.g. NVIDIA Nemotron 3 Ultra, 550B, ~190GB at 2-bit) are openWeights: true but runsLocally: false because no consumer tier fits them."
13
  },
14
- "updated": "2026-10-24",
15
  "counts": {
16
- "models": 169,
17
- "local": 126,
18
  "cloud": 43,
19
  "families": 32
20
  },
@@ -3057,6 +3057,63 @@
3057
  "runtimes": [],
3058
  "bestFor": "Agentic coding, Reasoning, Long context",
3059
  "ollamaCommand": null
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3060
  }
3061
  ]
3062
  }
 
11
  "note": "A model fits a device when its estimatedLoadGb is within the memory budget: ~70% of unified memory up to 32GB, scaling linearly to ~85% at 128GB and above (high-RAM Macs can wire more memory to the GPU). Q4_K_M needs roughly 0.6 GB per billion parameters.",
12
  "fieldSemantics": "runsLocally = a registry-verified Ollama build fits at least one consumer RAM tier tracked here (up to 256GB). openWeights = the weights are publicly downloadable. Some open-weight giants (e.g. NVIDIA Nemotron 3 Ultra, 550B, ~190GB at 2-bit) are openWeights: true but runsLocally: false because no consumer tier fits them."
13
  },
14
+ "updated": "2026-10-26",
15
  "counts": {
16
+ "models": 172,
17
+ "local": 129,
18
  "cloud": 43,
19
  "families": 32
20
  },
 
3057
  "runtimes": [],
3058
  "bestFor": "Agentic coding, Reasoning, Long context",
3059
  "ollamaCommand": null
3060
+ },
3061
+ {
3062
+ "model": "NVIDIA Nemotron 3.5 Bolt E4B",
3063
+ "family": "Nemotron",
3064
+ "params": 4,
3065
+ "quantization": "Q4_K_M",
3066
+ "minRamGb": 12,
3067
+ "estimatedLoadGb": 7.9,
3068
+ "kvKbPerToken": null,
3069
+ "runsLocally": true,
3070
+ "openWeights": true,
3071
+ "ggufDiy": false,
3072
+ "runtimes": [
3073
+ "ollama",
3074
+ "llama.cpp",
3075
+ "lm-studio"
3076
+ ],
3077
+ "bestFor": "Voice assistants, Tool calling, Chat",
3078
+ "ollamaCommand": "ollama run nemotron-3.5-bolt:e4b"
3079
+ },
3080
+ {
3081
+ "model": "Granite Switch 4.1 8B",
3082
+ "family": "Granite",
3083
+ "params": 8.8,
3084
+ "quantization": "Q4_K_M",
3085
+ "minRamGb": 12,
3086
+ "estimatedLoadGb": 6.4,
3087
+ "kvKbPerToken": null,
3088
+ "runsLocally": true,
3089
+ "openWeights": true,
3090
+ "ggufDiy": false,
3091
+ "runtimes": [
3092
+ "ollama",
3093
+ "llama.cpp",
3094
+ "lm-studio"
3095
+ ],
3096
+ "bestFor": "Chat, Decisions, Safety checks",
3097
+ "ollamaCommand": null
3098
+ },
3099
+ {
3100
+ "model": "Granite Switch 4.1 30B",
3101
+ "family": "Granite",
3102
+ "params": 30,
3103
+ "quantization": "Q4_K_M",
3104
+ "minRamGb": 32,
3105
+ "estimatedLoadGb": 21,
3106
+ "kvKbPerToken": null,
3107
+ "runsLocally": true,
3108
+ "openWeights": true,
3109
+ "ggufDiy": false,
3110
+ "runtimes": [
3111
+ "ollama",
3112
+ "llama.cpp",
3113
+ "lm-studio"
3114
+ ],
3115
+ "bestFor": "Chat, Decisions, Safety checks",
3116
+ "ollamaCommand": null
3117
  }
3118
  ]
3119
  }