self-hosted

Self-hostable open-weight models

Models with downloadable weights you can run yourself: Llama, Qwen, DeepSeek, Gemma, Phi, Mistral's open lines, and more.

What this is Ranked by capability match + real benchmark scores + live pricing, daily-refreshed. Benchmark sources: Aider Polyglot (coding) and Artificial Analysis Intelligence Index (overall, coding, agentic). Full methodology →

Download this catalog: CSV · JSON (regenerated 2026-09-27), or query it live via the API.

#ModelScoreIn / 1MOut / 1MContext
1 Qwen: Qwen3.8 Max Prime NEWqwen/qwen3.8-max-prime — $4.000 $12.000 1,000,000 Details →
2 Cohere: Command A+ NEWcohere/command-a-plus 17 1/5 $0.300 $1.500 192,000 Details →
3 Qwen: Qwen3.8 Omni Flash NEWqwen/qwen3.8-omni-flash — $0.150 $0.470 1,000,000 Details →
4 DeepSeek: DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash 66 1/5 $0.035 $0.290 1,048,576 Details →
5 DeepSeek: DeepSeek V4.1 Flash (batch)deepseek/deepseek-v4.1-flash:batch 66 1/5 $0.112 $0.336 1,048,576 Details →
6 Qwen: Qwen3.8 Max (0902)qwen/qwen3.8-max-0902 77 1/5 $2.000 $6.000 1,000,000 Details →
7 Qwen: Qwen3.8 Flashqwen/qwen3.8-flash — $0.150 $0.470 1,000,000 Details →
8 DeepSeek: DeepSeek V4 Flash Vision Expdeepseek/deepseek-v4-flash-vision-exp 58 1/5 $0.216 $0.647 1,048,576 Details →
9 Qwen: Qwen3.8 27Bqwen/qwen3.8-27b 56 1/5 $0.420 $3.000 1,000,000 Details →
10 Qwen: Qwen3.8 27B (free)qwen/qwen3.8-27b:free 56 1/5 Free Free 262,144 Details →
11 Qwen: Qwen3.7 Flashqwen/qwen3.7-flash — $0.030 $0.130 1,000,000 Details →
12 Qwen: Qwen3.7 Plusqwen/qwen3.7-plus 40 1/5 $0.320 $1.280 1,000,000 Details →
13 NVIDIA: Nemotron 3 Nano Omni (free)nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free — Free Free 256,000 Details →
14 Qwen: Qwen3.5 Plus 2026-04-20qwen/qwen3.5-plus-20260420 — $0.300 $1.800 1,000,000 Details →
15 Qwen: Qwen3.6 Flashqwen/qwen3.6-flash — $0.188 $1.125 1,000,000 Details →
16 Qwen: Qwen3.6 35B A3Bqwen/qwen3.6-35b-a3b 27 1/5 $0.150 $1.000 262,144 Details →
17 Qwen: Qwen3.6 27Bqwen/qwen3.6-27b 33 1/5 $0.320 $3.200 262,144 Details →
18 Google: Gemma 4 26B A4B google/gemma-4-26b-a4b-it 51 2/5 $0.068 $0.225 262,144 Details →
19 Google: Gemma 4 26B A4B (free)google/gemma-4-26b-a4b-it:free 51 2/5 Free Free 262,144 Details →
20 Google: Gemma 4 31Bgoogle/gemma-4-31b-it 54 2/5 $0.090 $0.340 262,144 Details →
21 Google: Gemma 4 31B (free)google/gemma-4-31b-it:free 54 2/5 Free Free 262,144 Details →
22 Qwen: Qwen3.6 Plusqwen/qwen3.6-plus 43 1/5 $0.325 $1.950 1,000,000 Details →
23 Mistral: Mistral Small 4mistralai/mistral-small-2603 4 1/5 $0.150 $0.600 262,144 Details →
24 Mistral: Mistral Small 4 (batch)mistralai/mistral-small-2603:batch 4 1/5 $0.075 $0.300 262,144 Details →
25 Qwen: Qwen3.5-9Bqwen/qwen3.5-9b 18 1/5 $0.100 $0.150 262,144 Details →
26 Qwen: Qwen3.5-35B-A3Bqwen/qwen3.5-35b-a3b 29 1/5 $0.312 $1.250 262,144 Details →
27 Qwen: Qwen3.5-27Bqwen/qwen3.5-27b 36 1/5 $0.195 $1.560 262,144 Details →
28 Qwen: Qwen3.5-122B-A10Bqwen/qwen3.5-122b-a10b 22 1/5 $0.260 $2.080 262,144 Details →
29 Qwen: Qwen3.5-Flashqwen/qwen3.5-flash-02-23 — $0.065 $0.260 1,000,000 Details →
30 Qwen: Qwen3.5 Plus 2026-02-15qwen/qwen3.5-plus-02-15 — $0.260 $1.560 1,000,000 Details →
31 Qwen: Qwen3.5 397B A17Bqwen/qwen3.5-397b-a17b 53 2/5 $0.550 $3.500 262,144 Details →
32 Mistral: Mistral Large 3 2512mistralai/mistral-large-2512 — $0.500 $1.500 262,144 Details →
33 Mistral: Mistral Large 3 2512 (batch)mistralai/mistral-large-2512:batch — $0.250 $0.750 262,144 Details →
34 Qwen: Qwen3 VL 30B A3B Thinkingqwen/qwen3-vl-30b-a3b-thinking — $0.200 $2.400 262,144 Details →
35 NVIDIA: Nemotron 3.5 Lightningnvidia/nemotron-3.5-lightning 17 1/5 $0.080 $0.200 1,000,000 Details →
36 DeepSeek: DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash — $0.047 $0.094 1,048,576 Details →
37 NVIDIA: Nemotron 3 Nano 30B A3Bnvidia/nemotron-3-nano-30b-a3b 6 1/5 $0.050 $0.200 262,144 Details →
38 Mistral: Ministral 3 8B 2512 (batch)mistralai/ministral-8b-2512:batch — $0.075 $0.075 262,144 Details →
39 DeepSeek: DeepSeek V4 Flash 0731deepseek/deepseek-v4-flash-0731 — $0.021 $0.320 1,310,720 Details →
40 Mistral: Ministral 3 14B 2512mistralai/ministral-14b-2512 — $0.200 $0.200 262,144 Details →
41 Mistral: Ministral 3 8B 2512mistralai/ministral-8b-2512 — $0.150 $0.150 262,144 Details →
42 Mistral: Mistral Small 3.2 24Bmistralai/mistral-small-3.2-24b-instruct 60 1/5 $0.094 $0.250 256,000 Details →
43 Meta: Llama 4 Scoutmeta-llama/llama-4-scout 8 1/5 $0.100 $0.300 1,310,720 Details →
44 NVIDIA: Nemotron 3 Supernvidia/nemotron-3-super-120b-a12b 17 1/5 $0.080 $0.450 262,144 Details →
45 Qwen: Qwen3 VL 8B Instructqwen/qwen3-vl-8b-instruct 6 1/5 $0.117 $0.455 262,144 Details →
46 Qwen: Qwen3 VL 30B A3B Instructqwen/qwen3-vl-30b-a3b-instruct 8 1/5 $0.150 $0.600 262,144 Details →
47 Meta: Llama 4 Maverickmeta-llama/llama-4-maverick 13 2/5 $0.188 $0.652 1,048,576 Details →
48 DeepSeek: DeepSeek V4 Pro 0423deepseek/deepseek-v4-pro 75 2/5 $0.348 $0.696 1,048,576 Details →
49 Qwen: Qwen3 Next 80B A3B Thinkingqwen/qwen3-next-80b-a3b-thinking — $0.150 $1.200 262,144 Details →
50 Qwen: Qwen3 VL 235B A22B Instructqwen/qwen3-vl-235b-a22b-instruct 12 1/5 $0.210 $1.900 262,144 Details →

The Model Movers Report

One email every Friday, built from this site's own rankings: the current top five by benchmark score, every model released in the last seven days, and one note worked out from that week's numbers. You can unsubscribe from any issue with one click.