10 Qwen3.8 Max (2.4T-A95B) Alibaba 80±1 local ~380GB 262k llama.cpp › copy 13 Kimi K3 Moonshot AI 79±1 local ~884GB 1M llama.cpp › copy 16 MiMo-V2.6-Flash (309B-A15B) Xiaomi 78±1 * local ~136GB 1M llama.cpp › copy 17 Qwen3.8-Flash-Next Alibaba 78±1 local ~59GB 262k llama.cpp › copy 18 Hy4 preview (770B-A49B) Tencent 77±1 local ~248GB 1M llama.cpp › copy 21 GLM-5.3 Z.ai 77±1 * local ~239GB 1M llama.cpp › copy 22 DeepSeek V4.1 Flash DeepSeek 77±1 * local ~537GB 1M how to run › copy 23 MiMo-V2.6-Pro Xiaomi 77±1 * local ~443GB 1M llama.cpp › copy 24 GLM-5.3-Flash Z.ai 76±1 local ~103GB 1M llama.cpp › copy 26 GLM-5.2 Z.ai 75±1 local ~121GB 1M llama.cpp › copy +1 28 DeepSeek V4 Pro (0813) DeepSeek 75±1 local ~1010GB 1M llama.cpp › copy 31 Qwen3.8 27B Alibaba 75±1 local ~18GB 262k LM Studio › copy 33 DeepSeek V4 Flash (0731) DeepSeek 75±1 local ~92GB 1M llama.cpp › copy 50 Kimi K2.6 Moonshot AI 71±1 local ~317GB 262k llama.cpp › copy 51 Hunyuan Hy3 (295B-A21B) Tencent 71±1 local ~96GB 256k llama.cpp › copy 54 GLM-5.1 Z.ai 70±1 local ~119GB 200k llama.cpp › copy +1 59 MiniMax-M3 MiniMax 69±1 * local ~137GB 1M llama.cpp › copy 60 Kimi K2.7 Code Moonshot AI 69±1 * local ~359GB 262k llama.cpp › copy 61 MiMo-V2.5-Pro (1T-A42B) Xiaomi 69±1 local ~163GB 1M llama.cpp › copy 63 DeepSeek V4 Pro (Preview, Apr 2026) DeepSeek 69±1 local ~604GB llama.cpp › copy 64 Ornith 1.5 35B-A3B DeepReinforce 68±3 † local ~22GB 262k LM Studio › copy 65 Solar Open 2 (250B-A15B) Upstage 68±4 * † local ~527GB 1M how to run › copy 69 KAT-Coder-V2.5-Dev (35B-A3B) Kuaishou (Kwaipilot) 67±3 † local ~22GB 262k LM Studio › copy 71 MiMo-V2.5 (310B-A15B) Xiaomi 67±1 local ~100GB 1M llama.cpp › copy 75 GLM-5 Z.ai 67±1 local ~119GB 205k llama.cpp › copy 77 DeepSeek V4 Flash (Preview, Apr 2026) DeepSeek 67±1 local ~103GB llama.cpp › copy 78 DeepSeek V4 Flash Vision (Exp) DeepSeek 67±8 * † local ~104GB 1M llama.cpp › copy 79 Kimi K2.5 Moonshot AI 67±1 local ~160GB 262k llama.cpp › copy 80 GLM-4.7 Z.ai 67±1 local ~58GB 205k llama.cpp › copy 82 A.X K2 (688B-A33B) SK Telecom 66±4 * † local ~1454GB 262k how to run › copy 83 Llama 3.1 Nemotron Ultra 253B v1 NVIDIA 66±5 * local ~82GB 128k llama.cpp › copy 84 Qwen3.6 27B Alibaba 66±2 local ~18GB 262k LM Studio › copy 87 Inkling (975B-A41B) Thinking Machines 65±1 local ~156GB 1M llama.cpp › copy 88 Inkling Small (266B-A12B) Thinking Machines 65±1 local ~86GB 1M llama.cpp › copy 89 Qwen3.5 397B-A17B Alibaba 65±1 local ~65GB 262k llama.cpp › copy 92 Nemotron 3 Ultra 550B-A55B NVIDIA 65±2 local ~89GB 1M llama.cpp › copy 96 Qwen3 235B-A22B Thinking 2507 Alibaba 65±3 local ~76GB 131k llama.cpp › copy 97 MiniMax-M2.7 MiniMax 65±1 local ~75GB 205k llama.cpp › copy 98 Qwen3.6 35B-A3B Alibaba 64±1 local ~23GB 262k LM Studio › copy 101 K-EXAONE 2.0 (750B-A37B) LG AI Research 64±4 * † local ~791GB 262k how to run › copy 105 Step 3.5 Flash (196B-A11B) StepFun 64±5 * † local ~118GB 256k LM Studio › copy 106 MiniMax-M2.1 MiniMax 64±1 * local ~79GB llama.cpp › copy 107 MiniMax-M2.5 MiniMax 64±1 local ~75GB 205k llama.cpp › copy 108 Nemotron Cascade 2 30B-A3B NVIDIA 63±2 † local ~20GB 256k LM Studio › copy 110 Qwen3.5 27B Alibaba 63±1 local ~18GB 262k LM Studio › copy 113 Ornith 1.0 35B (35B-A3B) DeepReinforce 63±2 local ~23GB 262k LM Studio › copy 114 DeepSeek V3.1 Terminus DeepSeek 63±2 † local ~230GB 164k llama.cpp › copy 115 Qwen3.5 122B-A10B Alibaba 63±1 local ~75GB 262k LM Studio › copy 116 Muse Glimmer 30B Meta 63±1 local ~19GB 131k llama.cpp › copy 117 Hy3 preview (295B-A21B) Tencent 63±1 local ~629GB how to run › copy 118 Laguna XS 2.1 (33B-A3B) Poolside 63±3 † local ~21GB 262k LM Studio › copy 119 Gemma 4 31B Google 62±1 local ~21GB 262k LM Studio › copy +1 120 DeepSeek V3.1 DeepSeek 62±1 local ~110GB 131k llama.cpp › copy 124 DeepSeek V3.2 DeepSeek 62±1 local ~110GB 128k llama.cpp › copy 125 Laguna M.1 Poolside 62±1 local ~135GB 262k LM Studio › copy 126 GigaChat 3.5 Ultra Reasoning (432B-A28B) Sber 62±3 * † local ~461GB 262k how to run › copy 127 DeepSeek R1 0528 DeepSeek 62±1 local ~110GB 164k llama.cpp › copy +2 130 Nemotron 3 Super 120B-A12B NVIDIA 61±2 † local ~73GB 262k LM Studio › copy 131 GLM-4.6 Z.ai 61±1 local ~58GB 205k llama.cpp › copy 133 Gemma 4 26B-A4B Google 61±1 local ~18GB 262k LM Studio › copy 135 MiMo-V2.6-Distill-Qwen-9B Xiaomi 61±3 * † local ~7.8GB LM Studio › copy 136 Seed-OSS 36B Instruct ByteDance Seed 61±3 † local ~23GB 524k LM Studio › copy 138 Kimi K2 Thinking Moonshot AI 60±1 * local ~160GB 262k llama.cpp › copy 139 K-EXAONE (236B-A23B) LG AI Research 60±3 † local ~152GB 262k llama.cpp › copy 141 Nemotron 3.5 Lightning 30B-A3B NVIDIA 60±5 * local ~20GB 262k LM Studio › copy 142 Gemma 4 12B Google 60±3 local ~8.5GB 262k LM Studio › copy 143 Laguna XS.2 Poolside 60±1 local ~21GB 262k LM Studio › copy 144 Devstral Small 2 24B Mistral 59±4 * † local ~17GB 262k llama.cpp › copy 145 Qwen3-VL 235B-A22B Thinking Alibaba 59±6 * † local ~111GB 256k LM Studio › copy 146 Nemotron-Labs-3 Puzzle 75B-A9B NVIDIA 59±4 * † local ~35GB llama.cpp › copy 148 Kimi K2 Instruct (0711) Moonshot AI 59±1 local ~160GB 131k llama.cpp › copy 149 MiniMax-M2 MiniMax 59±1 * local ~75GB 205k llama.cpp › copy 150 DeepSeek V3.2 Exp DeepSeek 59±1 local ~726GB how to run › copy 152 MiMo-V2-Flash (309B-A15B) Xiaomi 59±1 * local ~108GB llama.cpp › copy 153 Spark-X2.5 4B iFLYTEK (SparkLLM) 59±2 † local ~3.9GB 1M LM Studio › copy 154 DeepSeek R1 (Jan '25) DeepSeek 59±1 local ~110GB 128k llama.cpp › copy +2 159 Qwen3-Next 80B-A3B Thinking Alibaba 58±4 * † local ~27GB 131k llama.cpp › copy 160 Qwen3.5 35B-A3B Alibaba 58±1 local ~23GB 262k LM Studio › copy 161 Qwen3-Coder 480B-A35B Instruct Alibaba 58±1 local ~78GB 262k llama.cpp › copy 163 Phi-4-reasoning 14B Microsoft 58±3 * † local ~10GB 33k LM Studio › copy 164 DeepSeek V3 0324 DeepSeek 58±1 local ~108GB 164k llama.cpp › copy 165 GLM-4.5 Z.ai 58±2 local ~58GB 131k llama.cpp › copy 167 Granite 4.2 30B IBM 58±2 † local ~19GB 131k LM Studio › copy 168 Ministral 3 8B Reasoning 2512 Mistral 58±6 * † local ~6.5GB 262k LM Studio › copy 169 Mistral Medium 3.5 (128B) Mistral 57±1 local ~42GB 262k llama.cpp › copy 170 Qwen3 235B-A22B (Apr 2025, hybrid) Alibaba 57±1 local ~76GB 131k llama.cpp › copy +1 172 DeepSeek V3 (Dec '24) DeepSeek 57±1 local ~337GB 131k llama.cpp › copy +1 173 A.X K1 (519B-A33B) SK Telecom 57±4 * † local ~1091GB how to run › copy 174 Command A+ (05-2026) Cohere 57±3 local ~71GB 128k llama.cpp › copy 175 INTELLECT-3 Prime Intellect 57±6 * † local ~48GB llama.cpp › copy 176 MiniCPM5 2B OpenBMB 57±2 † local ~3GB 131k LM Studio › copy 178 QwQ-32B-Preview Alibaba 57±4 * local ~21GB 33k LM Studio › copy 181 gpt-oss-120b OpenAI 57±1 local ~67GB 131k llama.cpp › copy 182 Qwen3 235B-A22B Instruct 2507 Alibaba 57±1 local ~76GB 262k llama.cpp › copy 183 LongCat-Flash-Chat (560B-A27B) Meituan 56±2 † local ~333GB 128k LM Studio › copy 184 Kimi K2 Instruct (0905) Moonshot AI 56±4 * local ~298GB llama.cpp › copy 185 GLM-4.5-Air Z.ai 56±3 local ~35GB 131k llama.cpp › copy 189 Nemotron-Cascade-14B-Thinking NVIDIA 56±1 † local ~10GB LM Studio › copy 190 Mellum2.1 12B-A2.5B Thinking JetBrains 56±3 † local ~8.3GB LM Studio › copy 191 gpt-oss-20b OpenAI 56±2 local ~14GB 131k llama.cpp › copy 192 Qwen3-Next 80B-A3B Instruct Alibaba 56±2 † local ~27GB 131k llama.cpp › copy 193 Trinity-Large-Thinking Arcee AI 56±1 * local ~141GB 512k llama.cpp › copy 194 Mistral Large 3 (675B-A41B) Mistral 55±1 * local ~108GB 262k llama.cpp › copy 196 Qwen3-VL 235B-A22B Instruct Alibaba 55±4 * † local ~134GB 256k LM Studio › copy 197 Phi-4-reasoning-plus 14B Microsoft 55±2 local ~10GB LM Studio › copy 198 Qwen3-Coder-Next (80B-A3B) Alibaba 55±3 * local ~27GB 262k llama.cpp › copy 201 Qwen3-VL 32B Thinking Alibaba 55±3 * † local ~22GB 256k LM Studio › copy 203 Apriel 1.6 15B Thinker ServiceNow 55±2 * † local ~10GB LM Studio › copy 204 OpenReasoning-Nemotron-32B NVIDIA 54±2 † local ~21GB LM Studio › copy 205 OpenThinker 32B Open Thoughts 54±9 * † local ~21GB 33k LM Studio › copy 206 Devstral 2 (2512) Mistral 54±1 local ~75GB 256k LM Studio › copy 207 Qwen3-VL 30B-A3B Thinking Alibaba 54±3 * † local ~21GB 256k LM Studio › copy 208 Qwen3.5 9B Alibaba 54±1 local ~7.4GB 262k llama.cpp › copy +3 209 Qwen3 30B-A3B Thinking 2507 Alibaba 54±3 local ~20GB 262k LM Studio › copy 210 Command A Reasoning (08-2025) Cohere 54±7 * † local ~34GB llama.cpp › copy 212 Qwen2.5 Coder 32B Instruct Alibaba 53±1 local ~21GB 131k LM Studio › copy +2 217 DeepSeek-V2.5 DeepSeek 53±1 local ~151GB 164k Ollama › copy 218 Qwen3 32B Alibaba 53±1 local ~20GB 131k LM Studio › copy 219 Granite 4.2 8B IBM 53±2 † local ~6.6GB 131k LM Studio › copy 221 DeepSeek-Coder-V2 Instruct (0724) DeepSeek 53±2 * local ~151GB llama.cpp › copy 222 Nemotron-Cascade-8B NVIDIA 53±2 † local ~6.3GB LM Studio › copy 223 Granite 4.1 8B IBM 53±1 local ~7GB 131k LM Studio › copy 224 Nemotron-Cascade-8B-Thinking NVIDIA 53±2 † local ~6.3GB LM Studio › copy 225 Ministral 3 3B Reasoning 2512 Mistral 52±6 * † local ~3.5GB 262k LM Studio › copy 227 Qwen2.5 32B Instruct Alibaba 52±1 local ~21GB 131k LM Studio › copy 228 Qwen3 30B-A3B Instruct 2507 Alibaba 52±1 † local ~20GB 262k LM Studio › copy 233 DeepSeek-V2 Chat (0628) DeepSeek 52±2 * local ~66GB llama.cpp › copy 234 Llama 3.3 Nemotron Super 49B v1.5 NVIDIA 52±2 † local ~18GB 131k llama.cpp › copy 235 DeepSeek-V2.5 (1210) DeepSeek 52±1 * local ~82GB llama.cpp › copy 236 Olmo 3.1 32B Think Ai2 52±3 † local ~21GB 66k LM Studio › copy 237 QwQ-32B Alibaba 52±1 local ~21GB 131k LM Studio › copy 239 Falcon-H1R 7B TII 52±4 * † local ~6GB 262k LM Studio › copy 241 Olmo 3 32B Think Ai2 51±3 † local ~21GB 66k LM Studio › copy 242 Granite 4.1 30B IBM 51±2 † local ~19GB Ollama › copy 243 s1.1 32B Simple Scaling 51±9 * local ~22GB 33k LM Studio › copy 244 Qwen3 Coder 30B-A3B Instruct Alibaba 51±2 local ~20GB 160k LM Studio › copy 247 Qwen3-VL 8B Thinking Alibaba 51±2 * † local ~6.8GB 256k llama.cpp › copy 250 Qwen2.5 Coder 14B Instruct Alibaba 51±1 local ~10GB 131k LM Studio › copy 251 DeepSeek-Coder-V2 Instruct (0614) DeepSeek 51±1 local ~92GB llama.cpp › copy 252 Mistral Large 2 (2411) Mistral 51±2 * local ~78GB 128k llama.cpp › copy 253 Llama 3.1 405B Instruct Meta 51±1 local ~160GB 128k llama.cpp › copy 254 Llama 4 Maverick (402B-A17B) Meta 51±2 local ~65GB 1M llama.cpp › copy 255 Qwen2.5 72B Instruct Alibaba 51±1 local ~30GB 33k llama.cpp › copy +2 257 Nemotron Nano 9B v2 NVIDIA 50±1 local ~7.6GB 131k llama.cpp › copy 260 Mistral Large 2 (2407) Mistral 50±1 local ~45GB 128k llama.cpp › copy 261 Command A (03-2025) Cohere 50±2 † local ~37GB 256k llama.cpp › copy 264 Hermes 3 Llama 3.1 405B Nous Research 50±2 * local ~114GB 131k llama.cpp › copy 265 Qwen3 4B Thinking 2507 Alibaba 50±6 * † local ~3.9GB 262k LM Studio › copy 267 Ministral 3 14B Reasoning 2512 Mistral 50±4 local ~10GB 262k llama.cpp › copy 268 Phi-4 14B Microsoft 50±1 local ~10GB 16k LM Studio › copy 269 Granite 4.2 3B IBM 50±2 local ~3.7GB 131k LM Studio › copy 270 Mellum2 12B-A2.5B Thinking JetBrains 50±2 † local ~8.6GB 131k LM Studio › copy 272 EXAONE 4.0 32B LG AI Research 49±3 * † local ~20GB 131k LM Studio › copy 273 Olmo 3 7B Think Ai2 49±2 local ~5.8GB 66k LM Studio › copy 274 Gemma 2 27B IT Google 49±2 local ~18GB 8k LM Studio › copy 276 InternLM2.5 20B Chat Shanghai AI Laboratory (InternLM) 49±2 local ~14GB 33k llama.cpp › copy 277 Qwen3.5 4B Alibaba 49±1 local ~4.3GB 262k llama.cpp › copy +1 278 Llama 3.3 70B Instruct Meta 49±1 local ~24GB 128k llama.cpp › copy 279 Mistral Small 3 24B (2501) Mistral 49±4 * local ~16GB 32k LM Studio › copy 280 GigaChat 3.5 Ultra Instruct (432B-A28B) Sber 49±3 * † local ~457GB 262k how to run › copy 281 DeepSeek R1 Distill Qwen 14B DeepSeek 49±2 local ~10GB 131k LM Studio › copy 282 Qwen2.5 14B Instruct Alibaba 49±1 local ~10GB 131k LM Studio › copy 283 Mistral Small Instruct 22B (2409) Mistral AI 48±2 local ~15GB 32k LM Studio › copy 285 Qwen3 30B-A3B (Apr 2025, hybrid) Alibaba 48±1 † local ~19GB 33k LM Studio › copy 286 CodeQwen1.5 7B Chat Alibaba 48±1 local ~6.4GB 64k llama.cpp › copy 287 OpenCoder-8B-Instruct INF (infly) 48±1 local ~6.4GB llama.cpp › copy 288 DeepSeek R1 Distill Llama 70B DeepSeek 48±2 local ~30GB 131k llama.cpp › copy 289 Llama 3.1 70B Instruct Meta 48±1 local ~27GB 128k llama.cpp › copy +1 290 Qwen3 8B Alibaba 48±2 local ~6.1GB 33k LM Studio › copy 292 Qwen2.5 Coder 7B Instruct Alibaba 48±1 local ~6GB 131k LM Studio › copy 293 DeepSeek R1 Distill Qwen 32B DeepSeek 48±2 local ~21GB 131k LM Studio › copy 295 DeepSeek R1 0528 Qwen3 8B DeepSeek 47±4 local ~6.3GB 128k LM Studio › copy 296 Qwen3-VL 4B Thinking Alibaba 47±2 * † local ~4.1GB 256k llama.cpp › copy 297 Nemotron 3 Nano 30B-A3B NVIDIA 47±3 † local ~20GB 262k LM Studio › copy 299 DeepSeek LLM 67B Chat DeepSeek 47±2 local ~31GB 4k llama.cpp › copy 300 OpenThinker3 7B Open Thoughts 47±2 local ~6GB 33k LM Studio › copy 302 Qwen2 72B Instruct Alibaba 47±1 local ~33GB 131k llama.cpp › copy 304 Phi-3 Medium 128K Instruct Microsoft 47±2 local ~9.8GB 128k LM Studio › copy 305 OpenCodeInterpreter-DS-33B M-A-P 47±1 local ~22GB LM Studio › copy 307 Gemma 4 E4B Google 47±1 local ~6.9GB 131k LM Studio › copy 308 DeepSeek-Coder 33B Instruct DeepSeek 46±1 local ~21GB 16k LM Studio › copy 309 Codestral 22B v0.1 Mistral AI 46±1 local ~15GB 32k LM Studio › copy 310 WizardCoder 33B V1.1 WizardLM 46±1 local ~22GB 16k llama.cpp › copy 311 Gemma 2 9B IT Google 46±2 local ~7GB 8k LM Studio › copy 313 Qwen2.5 7B Instruct Alibaba 46±1 local ~6GB 131k LM Studio › copy 314 Mistral NeMo Instruct (2407) Mistral 46±2 local ~8.7GB 128k LM Studio › copy 315 Llama 3.1 Nemotron 70B Instruct NVIDIA 46±2 * local ~43GB 128k LM Studio › copy 316 Llama 3 70B Instruct Meta 46±1 local ~24GB 8k llama.cpp › copy 317 OpenCodeInterpreter-DS-6.7B M-A-P 46±1 local ~5.8GB llama.cpp › copy 319 Mixtral 8x22B Instruct v0.1 Mistral 46±1 local ~56GB 64k llama.cpp › copy 320 speechless-codellama-34b-v2.0 uukuguy 46±1 local ~23GB llama.cpp › copy 321 Magicoder-S-DS 6.7B UIUC (ise-uiuc) 46±1 local ~5.8GB llama.cpp › copy 322 Artigenz-Coder-DS-6.7B Artigenz 46±1 local ~5.8GB llama.cpp › copy 325 Yi 1.5 9B Chat 01.AI 45±2 local ~6.8GB 4k LM Studio › copy 326 Nemotron 3 Nano 4B NVIDIA 45±2 local ~3.8GB 262k LM Studio › copy 327 Pixtral 12B (2409) Mistral 45±3 * local ~9.6GB llama.cpp › copy 329 Granite 4.1 3B IBM 45±2 † local ~3.7GB Ollama › copy 330 XwinCoder-34B Xwin-LM 45±1 local ~23GB llama.cpp › copy 331 StarChat2 15B v0.1 Hugging Face 45±1 local ~12GB 16k llama.cpp › copy 332 Code Millenials 34B Bud Ecosystem 45±1 local ~23GB llama.cpp › copy 333 WizardLM 70B V1.0 Microsoft 45±2 local ~32GB llama.cpp › copy 334 DeepSeek-Coder 6.7B Instruct DeepSeek 45±1 local ~5.8GB 16k llama.cpp › copy 335 AceReason-Nemotron-1.0-14B NVIDIA 45±2 † local ~10GB LM Studio › copy 336 WaveCoder-Ultra-6.7B Microsoft 45±1 local ~5.8GB llama.cpp › copy 337 DBRX Instruct Databricks 45±1 local ~85GB 32k Ollama › copy 338 Yi-Coder 9B Chat 01.AI 45±1 local ~6.8GB 128k LM Studio › copy 339 MiniCPM3-4B OpenBMB 45±2 * local ~4GB 32k LM Studio › copy 340 Gemma 3 27B Google 44±1 local ~18GB 131k LM Studio › copy 341 Ministral 3 14B Mistral 44±4 * local ~9.1GB 262k LM Studio › copy 342 OpenThinker2 7B Open Thoughts 44±9 * † local ~6.4GB 33k llama.cpp › copy 343 DeepSeek-Coder 7B Instruct v1.5 DeepSeek 44±1 local ~5.9GB llama.cpp › copy 345 Qwen3 4B Alibaba 44±2 local ~3.7GB 33k LM Studio › copy 346 Gemma 3 12B Google 44±2 local ~9.1GB 131k LM Studio › copy 347 Yi 1.5 34B Chat 01.AI 44±2 local ~22GB 4k LM Studio › copy 349 Qwen3-VL 32B Instruct Alibaba 44±2 * † local ~22GB 256k LM Studio › copy 350 OpenChat 3.5 0106 OpenChat 44±1 local ~6.1GB 8k llama.cpp › copy 351 Athene-70B Nexusflow 43±3 * local ~29GB llama.cpp › copy 352 Granite 34B Code Instruct (8K) IBM 43±1 local ~22GB Ollama › copy 353 Phind CodeLlama 34B v2 Phind 43±1 local ~23GB 16k Ollama › copy 354 WhiteRabbitNeo-33B-v1 WhiteRabbitNeo 43±1 local ~22GB llama.cpp › copy 355 Magicoder-S-CL-7B UIUC (ise-uiuc) 43±1 local ~5.8GB llama.cpp › copy 356 Qwen2.5 Coder 3B Instruct Alibaba 43±1 local ~3.3GB 33k LM Studio › copy 357 Code Llama 70B Instruct Meta 43±1 local ~28GB 16k llama.cpp › copy 358 Qwen3-VL 30B-A3B Instruct Alibaba 43±2 * † local ~21GB 256k LM Studio › copy 359 speechless-coder-ds-6.7b uukuguy 43±1 local ~5.8GB llama.cpp › copy 360 Spark-X2.5 1.7B iFLYTEK (SparkLLM) 43±3 † local ~2.7GB 1M llama.cpp › copy 361 WizardLM-2 8x22B Microsoft 43±2 * local ~91GB Ollama › copy 362 WizardCoder Python 34B V1.0 WizardLM 43±1 local ~23GB 16k Ollama › copy 363 Phi-4-mini 3.8B Microsoft 43±2 † local ~3.8GB 128k LM Studio › copy 364 Code Tulu 2 34B Allen Institute for AI 43±3 * † local ~72GB how to run › copy 365 AceReason-Nemotron-1.1-7B NVIDIA 42±2 † local ~6GB LM Studio › copy 367 DeepSeek-Coder-V2 Lite Instruct (16B-A2.4B) DeepSeek 42±2 local ~11GB 128k LM Studio › copy 368 Solar Open 100B (102B-A12B) Upstage 42±3 * † local ~41GB 131k llama.cpp › copy 369 Qwen3 4B Instruct 2507 Alibaba 42±2 local ~3.9GB 262k LM Studio › copy 373 Mellum2 12B-A2.5B Instruct JetBrains 42±4 * † local ~8.6GB 131k LM Studio › copy 374 Llama 3.1 8B Instruct Meta 42±1 local ~6.2GB 128k LM Studio › copy +2 375 Llama 3.2 11B Vision Instruct Meta 42±3 * local ~8.8GB 128k Ollama › copy 376 speechless-starcoder2-15b uukuguy 42±1 local ~12GB llama.cpp › copy 377 Granite 3.0 8B Instruct IBM 42±2 local ~6.6GB 4k llama.cpp › copy 378 Qwen2-VL 7B Instruct Alibaba 42±3 * local ~6.4GB LM Studio › copy 379 Phi-3-mini-128k-instruct (Apr 2024) Microsoft 42±2 local ~9.5GB how to run › copy 380 Qwen2.5 3B Instruct Alibaba 42±2 local ~3.3GB 33k LM Studio › copy 381 Qwen1.5 72B Chat Alibaba 41±1 local ~31GB 32k llama.cpp › copy 382 Phi-3-mini-4k-instruct (June 2024 update) Microsoft 41±3 * local ~3.7GB LM Studio › copy 383 Qwen3-VL 8B Instruct Alibaba 41±2 * † local ~6.8GB 256k llama.cpp › copy 384 StarCoder2-15B-Instruct-v0.1 BigCode 41±1 local ~11GB 16k Ollama › copy 385 Command R Plus (08-2024) Cohere 41±1 local ~63GB 128k LM Studio › copy 386 Granite 8B Code Instruct (4K) IBM 41±2 † local ~6.6GB llama.cpp › copy 387 Granite 20B Code Instruct (8K) IBM 41±2 † local ~14GB Ollama › copy 388 Phi-3.5-mini-instruct Microsoft 41±2 local ~3.7GB 128k LM Studio › copy 389 Falcon3 7B Instruct TII 41±2 local ~5.9GB 32k LM Studio › copy 390 Qwen3-VL 4B Instruct Alibaba 41±2 * † local ~4.1GB 256k llama.cpp › copy 391 Qwen1.5 110B Chat Alibaba 40±2 * local ~45GB 32k llama.cpp › copy 392 Code Llama 70B Python Meta 40±1 local ~28GB 16k llama.cpp › copy 393 Dolphin 2.6 Mixtral 8x7B Cognitive Computations 40±1 local ~29GB llama.cpp › copy 394 Command R (08-2024) Cohere 40±1 local ~21GB 128k LM Studio › copy 395 DeepSeek-Coder 1.3B Instruct DeepSeek 40±1 local ~2.4GB 16k Ollama › copy 396 Command R Plus (04-2024) Cohere 40±1 local ~43GB 128k llama.cpp › copy 397 GLM-4.7-Flash Z.ai 40±6 * local ~19GB 200k LM Studio › copy 398 Olmo 3 7B Instruct Ai2 39±2 local ~5.8GB 66k LM Studio › copy 399 OpenChat 3.5 OpenChat 39±3 * † local ~6.1GB 8k llama.cpp › copy 400 Llama 3 8B Instruct Meta 39±1 local ~6.6GB 8k llama.cpp › copy 401 Code Llama 70B (base) Meta 39±1 local ~28GB 16k llama.cpp › copy 402 Qwen2 7B Instruct Alibaba 39±3 * local ~6GB 131k LM Studio › copy 403 Falcon3 10B Instruct TII 39±2 local ~7.6GB 32k LM Studio › copy 404 Code Tulu 2 13B Allen Institute for AI 39±2 * † local ~29GB how to run › copy 405 WizardCoder Python 13B V1.0 WizardLM 39±3 local ~9.8GB 16k llama.cpp › copy 406 Qwen2.5 Coder 1.5B Instruct Alibaba 39±1 local ~2.4GB 33k LM Studio › copy 407 Gemma 4 E2B Google 39±2 local ~5.1GB 131k llama.cpp › copy 408 CodeGemma 7B IT (v1.0) Google 39±1 local ~7.1GB llama.cpp › copy 409 LFM2.5 2.6B Liquid AI 38±2 local ~3.2GB 131k LM Studio › copy 410 Granite 34B Code Base (8K) IBM 38±1 † local ~24GB llama.cpp › copy 412 speechless-starcoder2-7b uukuguy 38±1 local ~6.1GB llama.cpp › copy 414 Code Llama 34B Python Meta 38±1 local ~23GB 16k Ollama › copy 415 WizardCoder-15B-V1.0 WizardLM 38±1 local ~12GB llama.cpp › copy 416 Qwen2.5 1.5B Instruct Alibaba 37±2 local ~2.4GB 33k LM Studio › copy 417 Hermes 3 Llama 3.1 8B Nous Research 37±2 * local ~6.6GB 128k llama.cpp › copy 418 speechless-coding-7b-16k-tora uukuguy 37±1 local ~16GB how to run › copy 419 Qwen3 1.7B Alibaba 37±2 * local ~2.6GB 33k LM Studio › copy 420 OpenHermes 2.5 Mistral 7B Nous Research 37±3 * local ~6.1GB llama.cpp › copy 421 Gemma 3 4B Google 37±2 local ~3.8GB 131k LM Studio › copy 422 Code Llama 34B Instruct Meta 37±1 local ~22GB 16k LM Studio › copy 423 Command R7B (12-2024) Cohere 37±2 † local ~6.2GB 128k LM Studio › copy 424 A OpenHermes-2.5-Code-290k-13B ajibawa-2023 37±1 local ~9.8GB llama.cpp › copy 425 WizardLM 13B V1.2 Microsoft 36±3 local ~9.8GB llama.cpp › copy 426 Qwen1.5 14B Chat Alibaba 36±1 * local ~11GB 32k Ollama › copy 427 Granite 3.0 2B Instruct IBM 36±2 local ~3.2GB 4k llama.cpp › copy 428 DeepSeek-Coder 33B Base DeepSeek 36±1 local ~22GB 16k llama.cpp › copy 429 Qwen3-VL 2B Thinking Alibaba 36±3 * † local ~2.7GB 256k llama.cpp › copy 430 CodeQwen1.5 7B (base) Alibaba 36±1 local ~6.4GB 66k llama.cpp › copy 431 Code Llama 34B (base) Meta 36±1 local ~23GB 16k Ollama › copy 432 Llama 3.2 3B Instruct Meta 36±1 local ~3.4GB 131k LM Studio › copy 433 Granite 3B Code Instruct (2K) IBM 36±2 † local ~3.7GB llama.cpp › copy 434 Phi-2 (2.7B) Microsoft 35±1 local ~3.2GB 2k LM Studio › copy 435 Falcon3 3B Instruct TII 35±3 local ~3.4GB 32k LM Studio › copy 436 SmolLM3 3B Hugging Face 35±6 * † local ~3.3GB 66k LM Studio › copy 437 Tulu 2 DPO 70B Ai2 35±2 * † local ~32GB llama.cpp › copy 438 Qwen-14B-Chat Alibaba 35±1 * local ~11GB llama.cpp › copy 439 Code Llama 13B Instruct Meta 35±1 local ~9.8GB 16k llama.cpp › copy 440 Code Tulu 2 7B Allen Institute for AI 35±2 * † local ~16GB how to run › copy 441 LFM2.5 8B-A1B Liquid AI 35±2 local ~6.5GB 128k LM Studio › copy 442 DeepSeek-V2 Lite Chat (16B-A2.4B) DeepSeek 35±2 local ~11GB 32k LM Studio › copy 443 Tulu 2 70B Allen Institute for AI 35±2 * † local ~28GB llama.cpp › copy 444 Granite 8B Code Base (4K) IBM 35±1 † local ~6.6GB llama.cpp › copy 445 DeepSeek-Coder 6.7B Base DeepSeek 35±1 local ~5.8GB 16k llama.cpp › copy 446 Mixtral 8x7B Instruct v0.1 Mistral 34±1 local ~29GB 32k LM Studio › copy 447 MistralHermes-CodePro-7B-v1 beowolx 34±1 local ~6.1GB llama.cpp › copy 448 speechless-code-mistral-7b-v1.0 uukuguy 34±1 local ~17GB how to run › copy 449 Code Llama 13B Python Meta 34±1 local ~9.8GB 16k llama.cpp › copy 450 StarCoder2 15B BigCode 34±1 local ~11GB 16k LM Studio › copy 451 Mixtral 8x7B v0.1 (base) Mistral AI 34±1 local ~29GB 32k llama.cpp › copy 452 Xwin-LM 70B v0.1 Xwin-LM 34±2 * local ~28GB llama.cpp › copy 453 CodeGemma 7B (v1.0) Google 34±1 local ~7.1GB llama.cpp › copy 454 Qwen1.5 7B Chat Alibaba 34±1 * local ~6.5GB 32k llama.cpp › copy 455 EXAONE 3.5 2.4B Instruct LG AI Research 34±3 local ~3GB 33k LM Studio › copy 456 Granite 20B Code Base (8K) IBM 33±2 † local ~15GB llama.cpp › copy 457 Code Llama 7B Instruct Meta 33±1 local ~5.5GB 16k LM Studio › copy 458 Code Llama 7B Python Meta 33±1 local ~5.8GB 16k llama.cpp › copy 459 Code Llama 13B (base) Meta 33±1 local ~9.2GB 16k LM Studio › copy 460 SOLAR 10.7B Instruct v1.0 Upstage 33±1 local ~8.3GB 4k llama.cpp › copy 461 Qwen2.5 Coder 0.5B Instruct Alibaba 33±1 local ~1.8GB 33k LM Studio › copy 462 Gemma 1.1 7B IT Google 32±1 local ~7.1GB 8k llama.cpp › copy 463 Gemma 2 2B IT Google 32±2 local ~3.1GB 8k LM Studio › copy 464 DeepSeek LLM 7B Chat DeepSeek 32±3 * † local ~5.7GB Ollama › copy 465 SmolLM2 1.7B Instruct Hugging Face 32±4 * † local ~2.6GB 8k LM Studio › copy 466 Mistral 7B Instruct v0.2 Mistral 32±1 local ~6GB 32k LM Studio › copy 467 Llama 3.2 1B Instruct Meta 31±2 local ~2.2GB 131k LM Studio › copy 468 xDAN-L1-Chat-RL-v1 xDAN-AI 31±1 local ~6.1GB llama.cpp › copy 469 Granite 3.0 3B-A800M Instruct IBM 31±2 * † local ~3.7GB 4k llama.cpp › copy 470 Code Llama 7B (base) Meta 31±1 local ~5.5GB 16k LM Studio › copy +1 471 Llama 2 70B (base) Meta 30±1 local ~32GB 4k llama.cpp › copy 472 Granite 3B Code Base (2K) IBM 30±2 † local ~3.7GB llama.cpp › copy 473 StarCoder2 7B BigCode 30±1 local ~6.1GB 16k LM Studio › copy 474 Qwen3-VL 2B Instruct Alibaba 30±3 * † local ~2.7GB 256k llama.cpp › copy 475 StarCoder 15.5B BigCode 30±1 local ~11GB Ollama › copy 476 Gemma 7B (base) Google 30±1 local ~7.1GB 8k llama.cpp › copy 477 Llama 3 8B (base) Meta 30±1 local ~6.6GB llama.cpp › copy 478 StarCoder2 3B BigCode 29±1 local ~3.4GB 16k llama.cpp › copy 479 StarCoderBase 15.5B BigCode 29±1 local ~12GB llama.cpp › copy 480 Mistral 7B Instruct v0.1 Mistral 29±1 * local ~6.1GB 8k llama.cpp › copy 481 Llama 2 70B Chat Meta 28±2 * local ~32GB 4k llama.cpp › copy 483 DeepSeek-Coder 1.3B Base DeepSeek 28±1 local ~2.4GB 16k Ollama › copy 484 Stable Code 3B Stability AI 28±1 local ~3.3GB 16k llama.cpp › copy 485 OpenThinker 7B Open Thoughts 28±8 * † local ~6GB 33k LM Studio › copy 486 Mistral 7B v0.1 (base) Mistral AI 28±1 local ~5.8GB 8k LM Studio › copy 487 Gemma 7B IT Google 28±1 local ~7.1GB 8k llama.cpp › copy 488 Zephyr 7B beta Hugging Face 28±1 local ~6.1GB llama.cpp › copy 489 DeepSeek R1 Distill Qwen 7B DeepSeek 27±3 local ~6GB 131k LM Studio › copy 490 DeepSeek R1 Distill Llama 8B DeepSeek 27±4 local ~6.2GB 131k LM Studio › copy 491 Tulu 2 13B Allen Institute for AI 27±2 * † local ~9.8GB llama.cpp › copy 492 Tulu 2 DPO 13B Allen Institute for AI 27±2 * † local ~9.8GB llama.cpp › copy 493 Granite 3.0 1B-A400M Instruct IBM 27±2 * † local ~2.3GB 4k llama.cpp › copy 494 CodeGemma 2B (v1.0) Google 27±1 local ~3.2GB llama.cpp › copy 496 Falcon 40B (base) TII 26±2 * local ~27GB 2k Ollama › copy 497 StarCoderBase 7B BigCode 26±1 local ~6GB Ollama › copy 498 Falcon3 1B Instruct TII 26±4 local ~2.4GB 8k LM Studio › copy 499 Gemma 3 1B Google 25±2 † local ~2.2GB 32k LM Studio › copy 500 Gemma 2B (base) Google 25±1 local ~3.2GB 8k llama.cpp › copy 501 Qwen3.5 2B Alibaba 24±3 local ~2.9GB 262k llama.cpp › copy 502 Gemma 1.1 2B IT Google 24±1 local ~3.2GB 8k llama.cpp › copy 503 Tulu 2 DPO 7B Allen Institute for AI 23±2 * † local ~5.8GB llama.cpp › copy 504 StarCoderBase 3B BigCode 23±1 local ~3.4GB Ollama › copy 506 Llama 2 13B (base) Meta 22±1 local ~9.3GB 4k Ollama › copy 507 Tulu 2 7B Allen Institute for AI 22±2 * † local ~5.8GB llama.cpp › copy 508 Gemma 2B IT Google 22±1 local ~3.1GB 8k llama.cpp › copy 510 Llama 2 13B Chat Meta 21±2 * local ~9.8GB 4k llama.cpp › copy 512 OpenAssistant SFT-4 Pythia 12B LAION (OpenAssistant) 20±2 * local ~9.5GB llama.cpp › copy 513 Phi-1 (1.3B) Microsoft 19±5 * local ~2.4GB 2k llama.cpp › copy 514 GPT-J 6B EleutherAI 19±1 * local ~14GB 2k how to run › copy 515 Llama 2 7B (base) Meta 19±1 local ~5.5GB 4k Ollama › copy 517 Qwen3 0.6B Alibaba 17±3 * local ~1.8GB 33k LM Studio › copy 518 Llama 2 7B Chat Meta 16±2 * local ~5.8GB 4k llama.cpp › copy 519 Molmo 7B-D (0924) Ai2 14±5 * local ~35GB how to run › copy 520 GPT-Neo 2.7B EleutherAI 14±1 * local ~13GB 2k how to run › copy 521 DeepSeek R1 Distill Qwen 1.5B DeepSeek 14±5 local ~2.6GB 131k LM Studio › copy 522 ChatGLM-6B Zhipu AI (THUDM) 12±3 * local ~16GB how to run › copy 523 StableLM-Base-Alpha 7B Stability AI 5±2 local ~35GB 4k how to run › copy 524 Zyte-1B venkycs (aihub-app) 3±2 local ~2.2GB llama.cpp › copy — BLOOM 1.1B BigScience not yet calibrated local ~2.6GB 2k llama.cpp › copy — BLOOM 1.7B BigScience not yet calibrated local ~3.1GB 2k llama.cpp › copy — BLOOM 3B BigScience not yet calibrated local ~4GB 2k llama.cpp › copy — BLOOM 560M BigScience not yet calibrated local ~2.1GB 2k llama.cpp › copy — BLOOM 7.1B BigScience not yet calibrated local ~7.1GB 2k llama.cpp › copy — ERNIE 4.5 300B-A47B Baidu not yet calibrated local ~97GB 131k llama.cpp › copy — EXAONE Deep 2.4B LG AI Research not yet calibrated local ~3.2GB 33k Ollama › copy — EXAONE Deep 32B LG AI Research not yet calibrated local ~22GB 33k Ollama › copy — EXAONE Deep 7.8B LG AI Research not yet calibrated local ~6.5GB 33k llama.cpp › copy — Gemma 3n E2B IT Google not yet calibrated local ~4.7GB 32k llama.cpp › copy — Gemma 3n E4B Google not yet calibrated local ~6.2GB 32k llama.cpp › copy — Granite 4.0 1B IBM not yet calibrated local ~2.4GB 128k LM Studio › copy — Granite 4.0 350M IBM not yet calibrated local ~1.7GB 32k LM Studio › copy — Granite 4.0 H 1B IBM not yet calibrated local ~2.3GB 131k LM Studio › copy — Granite 4.0 H 350M IBM not yet calibrated local ~1.7GB 33k LM Studio › copy — Granite 4.0 H Micro 3B IBM not yet calibrated local ~3.4GB 131k LM Studio › copy — Granite 4.0 H Small (32B-A9B) IBM not yet calibrated local ~21GB 131k LM Studio › copy — Granite 4.0 H Tiny (7B-A1B) IBM not yet calibrated local ~5.6GB 131k LM Studio › copy — Granite 4.0 Micro 3B IBM not yet calibrated local ~3.5GB 131k LM Studio › copy — Jamba 1.5 Large AI21 Labs not yet calibrated local ~839GB how to run › copy — Jamba 1.5 Mini AI21 Labs not yet calibrated local ~34GB llama.cpp › copy — K2 Horizon 7B Institute of Foundation Models (MBZUAI) not scored yet local ~7.7GB 524k llama.cpp › copy — Ling 3.0 Flash (124B-A5B) inclusionAI (Ant Group) not yet calibrated local ~75GB 131k LM Studio › copy — Llama 3.3 Nemotron Super 49B v1 NVIDIA not scored yet local ~22GB llama.cpp › copy — LongCat 2.0 (1.6T-A48B) Meituan not scored yet local ~2155GB 1M how to run › copy — Magistral Small 1.0 (2506) Mistral not yet calibrated local ~16GB 128k LM Studio › copy — Magistral Small 1.1 (2507) Mistral not yet calibrated local ~16GB 40k LM Studio › copy — Magistral Small 1.2 Mistral not yet calibrated local ~16GB 131k LM Studio › copy — MiniMax-M1 (80k) MiniMax not yet calibrated local ~959GB 1M how to run › copy — Ministral 8B (2410) Mistral not yet calibrated local ~6.2GB 128k LM Studio › copy — Mistral 7B Instruct v0.3 Mistral not yet calibrated local ~5.8GB 32k LM Studio › copy — Mistral Small 3.1 24B (2503) Mistral not scored yet local ~17GB 128k llama.cpp › copy — Mistral Small 3.2 24B Mistral not scored yet local ~16GB 128k LM Studio › copy — Mistral Small 4 (119B-A6.5B) Mistral not yet calibrated local ~40GB 262k llama.cpp › copy — OLMo 7B Allen Institute for AI not yet calibrated local ~30GB 2k how to run › copy — OLMo 7B Instruct Ai2 not yet calibrated local ~16GB how to run › copy — Qwen2 0.5B Instruct Alibaba not yet calibrated local ~1.8GB 32k LM Studio › copy — Qwen2 1.5B Instruct Alibaba not yet calibrated local ~2.4GB 32k LM Studio › copy — Step 3.7 Flash (198B-A11B) StepFun not scored yet local ~104GB 256k llama.cpp › copy 
