Comparisons of Large Open Source AI Models (>150B)
Open source AI models with over 150 billion parameters. Models are considered Open Source (also commonly referred to as open weights) where their weights are accessible to download. This allows self-hosting on your own infrastructure and enables customizing the model such as through fine-tuning. Click on any model to see detailed metrics. For more details including relating to our methodology, see our FAQs.
DeepSeek V3.1 Terminus and 
Qwen3 235B A22B 2507 are the highest intelligence Large open source models, defined as those with >150B parameters, followed by DeepSeek V3.2 Exp & 
GLM-4.6.
Highlights
Intelligence
Artificial Analysis Intelligence Index; Higher is better
Estimate (independent evaluation forthcoming)
Total Parameters
Trainable parameters in billions
Further details
| Weights | Provider Benchmarks | |||||||
|---|---|---|---|---|---|---|---|---|
DeepSeek V3.1 Terminus (Reasoning) DeepSeek | 58 | 685B (37B active at inference time) | 128k | $1.9 | - | 🤗 |   | View |
Qwen3 235B A22B 2507 (Reasoning) Alibaba | 57 | 235B (22B active at inference time) | 256k | $2.6 | 86 | 🤗 |    +6 more | View |
DeepSeek V3.2 Exp (Reasoning) DeepSeek | 57 | 685B (37.0B active at inference time) | 128k | $0.3 | 26 | 🤗 | | View |
GLM-4.6 (Reasoning) Z AI | 56 | 357B (32B active at inference time) | 200k | $1.0 | 50 | 🤗 |  +1 more | View |
DeepSeek R1 0528 (May '25) DeepSeek | 52 | 685B (37B active at inference time) | 128k | $1.6 | - | 🤗 |   +12 more | View |
 Kimi K2 0905 Moonshot AI | 50 | 1.0KB (32.0B active at inference time) | 256k | $1.3 | 53 | 🤗 |  +3 more | View |
DeepSeek V3.2 Exp (Non-reasoning) DeepSeek | 46 | 685B (37.0B active at inference time) | 128k | $0.3 | 24 | 🤗 | +1 more | View |
 MiniMax M1 80k MiniMax | 46 | 456B (45.9B active at inference time) | 1.00M | $0.8 | - | 🤗 | | View |
DeepSeek V3.1 Terminus (Non-reasoning) DeepSeek | 46 | 685B (37B active at inference time) | 128k | $0.5 | - | 🤗 |  +2 more | View |
Qwen3 235B A22B 2507 Instruct Alibaba | 45 | 235B (22.0B active at inference time) | 256k | $1.2 | 49 | 🤗 | +9 more | View |
GLM-4.6 (Non-reasoning) Z AI | 45 | 357B (32B active at inference time) | 200k | $1.0 | 50 | 🤗 | | View |
Qwen3 Coder 480B A35B Instruct Alibaba | 42 | 480B (35.0B active at inference time) | 262k | $3.0 | 48 | 🤗 | +10 more | View |
MiniMax M1 40k MiniMax | 42 | 456B (45.9B active at inference time) | 1.00M | $0.8 | - | 🤗 | | View |
 Hermes 4 - Llama-3.1 405B (Reasoning) Nous Research | 42 | 406B | 128k | $1.5 | 37 | 🤗 | | View |
 Llama 3.1 Nemotron Ultra 253B v1 (Reasoning) NVIDIA | 38 | 253B | 128k | $0.9 | 38 | 🤗 | | View |
 Llama 4 Maverick Meta | 36 | 402B (17B active at inference time) | 1.00M | $0.4 | 142 | 🤗 | +12 more | View |
Hermes 4 - Llama-3.1 405B (Non-reasoning) Nous Research | 33 | 406B | 128k | $1.5 | 34 | 🤗 | | View |
Llama 3.1 Instruct 405B Meta | 28 | 405B | 128k | $3.5 | 34 | 🤗 | ?  +8 more | View |
MiniMax-Text-01 MiniMax | 26 | 456B (45.9B active at inference time) | 4.00M | $0.4 | - | 🤗 | | View |
 Jamba 1.7 Large AI21 Labs | 21 | 398B (94.0B active at inference time) | 256k | $3.5 | 44 | 🤗 | | View |
 R1 1776 Perplexity | 19 | 671B (37B active at inference time) | 128k | $3.5 | - | 🤗 | | View |