| Rank | Model | Provider | Score | Benchmarks | Inference | Agentic | Programming | Value | Price |
|---|---|---|---|---|---|---|---|---|---|
| 121 | DeepSeek-V3.2 (Thinking) deepseek-reasoner codeprogrammingtool use | DeepSeek | 12.0 Agentic | 47.7 | 0.0 | 12.0 | 42.5 | 0.0 | N/A |
| 122 | DeepSeek-V3.2 deepseek-v3.2 codeprogrammingtool use | DeepSeek | 12.0 Agentic | 54.0 | 22.1 | 12.0 | 42.5 | 75.5 | $0.26 in / $0.38 out |
| 123 | o3-mini o3-mini codeprogrammingtool use | OpenAI | 11.9 Agentic | 24.3 | 0.0 | 11.9 | 11.0 | 0.0 | N/A |
| 124 | Grok 4 Fast grok-4-fast multimodalvisionmulti-input reasoning | xAI | 11.8 Agentic | 54.0 | 0.0 | 11.8 | 0.0 | 0.0 | N/A |
| 125 | Qwen3 Max Thinking qwen3-max-thinking multimodalvisionmulti-input reasoning | Alibaba Cloud / Qwen Team | 11.2 Agentic | 45.6 | 30.9 | 11.2 | 51.5 | 25.0 | $1.2 in / $6 out |
| 126 | Nemotron 3 Ultra (550B A55B) nemotron-3-ultra-550b-a55b codeprogrammingtool use | NVIDIA | 10.7 Agentic | 52.9 | 41.8 | 10.7 | 40.6 | 44.5 | $0.5 in / $2.2 out |
| 127 | Qwen3.6-35B-A3B qwen3.6-35b-a3b multimodalvisionmulti-input reasoning | Alibaba Cloud / Qwen Team | 9.6 Agentic | 49.1 | 41.8 | 9.6 | 27.1 | 65.9 | $0.1 in / $0.95 out |
| 128 | GPT-4.1 mini gpt-4.1-mini-2025-04-14 multimodalvisionmulti-input reasoning | OpenAI | 8.9 Agentic | 17.8 | 84.1 | 8.9 | 1.9 | 57.0 | |
| 129 | GLM-4.7-Flash glm-4.7-flash codeprogrammingtool use | Zhipu AI | 8.6 Agentic | 35.2 | 0.0 | 8.6 | 17.4 | 0.0 | N/A |
| 130 | Nemotron 3 Super (120B A12B) nemotron-3-super-120b-a12b codeprogrammingtool use | NVIDIA | 7.2 Agentic | 43.5 | 41.8 | 7.2 | 22.2 | 81.9 | $0.085 in / $0.4 out |
| 131 | GPT-5.4 nano gpt-5.4-nano multimodalvisionmulti-input reasoning | OpenAI | 6.9 Agentic | 38.1 | 52.5 | 6.9 | 13.3 | 57.1 | $0.2 in / $1.25 out |
| 132 | Nemotron 3.5 Lightning (30B A3B) nemotron-3.5-lightning-30b-a3b codeprogrammingtool use | NVIDIA | 6.4 Agentic | 26.3 | 41.8 | 6.4 | 8.2 | 89.0 | $0.08 in / $0.2 out |
| 133 | GPT OSS 20B gpt-oss-20b textinference | OpenAI | 6.0 Agentic | 22.0 | 15.0 | 6.0 | 0.0 | 96.2 | $0.03 in / $0.14 out |
| 134 | Kimi K2-Instruct-0905 kimi-k2-instruct-0905 codeprogrammingtool use | Moonshot AI | 6.0 Agentic | 22.4 | 0.0 | 6.0 | 16.9 | 0.0 | |
| 135 | Sarvam-30B sarvam-30b codeprogrammingtool use | Sarvam AI | 5.4 Agentic | 44.3 | 0.0 | 5.4 | 3.9 | 0.0 | N/A |
| 136 | DeepSeek-V3.2-Speciale deepseek-v3.2-speciale codeprogrammingtool use | DeepSeek | 5.3 Agentic | 48.8 | 0.0 | 5.3 | 42.5 | 0.0 | |
| 137 | Qwen2.5 VL 72B Instruct qwen2.5-vl-72b multimodalvisionmulti-input reasoning | Alibaba Cloud / Qwen Team | 4.7 Agentic | 21.8 | 0.0 | 4.7 | 0.0 | 0.0 | N/A |
| 138 | Claude 3.5 Haiku claude-3-5-haiku-20241022 codeprogrammingtool use | Anthropic | 3.0 Agentic | 9.1 | 0.0 | 3.0 | 5.8 | 0.0 | |
| 139 | Nemotron 3 Nano (30B A3B) nemotron-3-nano-30b-a3b codeprogrammingtool use | NVIDIA | 3.0 Agentic | 42.1 | 41.8 | 3.0 | 5.3 | 91.5 | $0.05 in / $0.2 out |
| 140 | Qwen2.5 VL 32B Instruct qwen2.5-vl-32b multimodalvisionmulti-input reasoning | Alibaba Cloud / Qwen Team | 1.3 Agentic | 18.4 | 0.0 | 1.3 | 0.0 | 0.0 | N/A |
121
DeepSeek-V3.2 (Thinking)
DeepSeek
12.0
N/A
122
DeepSeek-V3.2
DeepSeek
12.0
$0.26 in / $0.38 out
123
o3-mini
OpenAI
11.9
N/A
124