open-source model leaderboards
This page is built for high-intent searches like “open LLM leaderboard” and “open source LLM leaderboard”, giving tracked open language models a dedicated ranking and comparison surface.
154
Tracked open LLMs
30
Providers
24
Free endpoints
8
Recent open launches
Top 20 Open-Source Models by Composite Score
Ranked by benchmark performance, capabilities, and context window
Current Top 25
| # | Model | Provider | Score | MMLU | Arena | Price | Context |
|---|---|---|---|---|---|---|---|
| #58 | DeepSeek V4 Pro | DeepSeek | 87 | - | 1463 | $1.17/M | 1.0M |
| #72 | GLM 5.2 | Zhipu AI | 83 | - | 1457 | $0.500/M | 1.0M |
| #73 | GLM 5.2 (batch) | Zhipu AI | 83 | - | 1457 | $0.700/M | 512K |
| #76 | GLM 5.1 | Zhipu AI | 82 | - | 1467 | $1.40/M | 205K |
| #78 | GLM 5 | Zhipu AI | 82 | - | 1457 | $0.950/M | 205K |
| #84 | DeepSeek V3.2 | DeepSeek | 81 | 88.5 | 1425 | $0.269/M | 164K |
| #87 | Gemma 4 31B | 81 | - | 1451 | $0.100/M | 262K | |
| #88 | Gemma 4 31B (free) | 81 | - | 1451 | Free | 262K | |
| #92 | Qwen3.5 397B A17B | Alibaba | 79 | - | 1442 | $0.500/M | 262K |
| #93 | R1 0528 | DeepSeek | 79 | 91.5 | 1380 | $0.500/M | 164K |
| #105 | MiniMax M2.7 | MiniMax | 78 | - | 1416 | $0.300/M | 205K |
| #106 | MiniMax M2.5 | MiniMax | 78 | - | 1390 | $0.220/M | 205K |
| #107 | Qwen3.5-122B-A10B | Alibaba | 78 | - | 1417 | $0.290/M | 262K |
| #108 | Gemma 2 27B | 77 | 75.2 | - | $0.650/M | 8K | |
| #109 | Qwen3.5-27B | Alibaba | 77 | - | 1408 | $0.195/M | 262K |
| #114 | MiMo-V2.5-Pro | Xiaomi | 76 | - | 1468 | $0.435/M | 1.1M |
| #115 | Qwen3.5-35B-A3B | Alibaba | 76 | - | 1395 | $0.140/M | 262K |
| #117 | Kimi K2.6 | Moonshot AI | 76 | - | 1460 | $0.950/M | 262K |
| #123 | Hy3 | Tencent | 75 | - | 1457 | $0.132/M | 262K |
| #130 | R1 | DeepSeek | 74 | 90.8 | 1369 | $0.700/M | 164K |
| #131 | MiniMax M3 | MiniMax | 74 | - | 1443 | $0.300/M | 1.0M |
| #132 | Inkling | thinkingmachines | 74 | - | 1442 | $0.950/M | 1.0M |
| #133 | MiniMax M3 (batch) | MiniMax | 74 | - | 1443 | $0.150/M | 524K |
| #134 | GLM 4.7 | Zhipu AI | 74 | - | 1442 | $0.400/M | 205K |
| #135 | GLM 4.6 | Zhipu AI | 74 | - | 1424 | $0.550/M | 205K |
Top Open Providers
Alibaba
32NVIDIA
13DeepSeek
12Zhipu AI
11Mistral AI
10Moonshot AI
8Meta
8Recent Open Releases
LFM2.5-2.6B (free)
Liquid AI
Aug 11, 2026 · Free
Nemotron 3.5 Lightning
NVIDIA
Aug 11, 2026 · $0.100/M
Nemotron 3.5 Lightning (free)
NVIDIA
Aug 11, 2026 · Free
Muse Glimmer 30B
meta
Aug 9, 2026 · $0.350/M
Ling 3.0 Tiny (free)
inclusionai
Aug 6, 2026 · Free
DeepSeek V4 Flash 0731
DeepSeek
Jul 31, 2026 · $0.080/M
Inkling Small
thinkingmachines
Jul 30, 2026 · $0.450/M
Ling-3.0-flash
inclusionai
Jul 23, 2026 · $0.021/M
Related Surfaces
This page is the tighter “open LLM leaderboard” landing page focused on rankings and table-based comparison, while the broader open-source models page is more explanatory and guide-oriented.
The ranking uses the live composite score, then surfaces benchmark columns, pricing, context window, and release recency so the open-source language model landscape is easier to compare.
Yes. This only includes coding/language models currently marked as open source in the live model catalog. It does not fill the ranking with demo entries.