| Rank | Model | Provider | Score | Benchmarks | Inference | Agentic | Programming | Value | Price |
|---|---|---|---|---|---|---|---|---|---|
| 41 | Ember-1 ember-1 codeprogrammingtool use | Fireworks AI | 60.9 overall | 0.0 | 74.3 | 0.0 | 74.6 | 9.1 | $3 in / $15 out |
| 42 | DeepSeek-V4.1-Flash-1017 deepseek-v4.1-flash-1017 multimodalvisionmulti-input reasoning | DeepSeek | 60.7 overall | 48.8 | 85.2 | 0.0 | 0.0 | 54.7 | |
| 43 | GPT-5.5 gpt-5.5 multimodalvisionmulti-input reasoning | OpenAI | 60.5 overall | 74.3 | 96.4 | 55.3 | 48.5 | 3.3 | $5 in / $30 out |
| 44 | GPT-5 High gpt-5-high-2025-08-07 multimodalvisionmulti-input reasoning | OpenAI | 60.5 overall | 60.5 | 0.0 | 0.0 | 0.0 | 0.0 | |
| 45 | Gemini 3.1 Pro gemini-3.1-pro-preview multimodalvisionmulti-input reasoning | Google | 59.6 overall | 70.6 | 85.2 | 46.7 | 61.7 | 14.0 | |
| 46 | Qwen3.7 Max qwen3.7-max multimodalvisionmulti-input reasoning | Alibaba Cloud / Qwen Team | 59.4 overall | 64.4 | 66.7 | 48.6 | 72.6 | 30.5 | $1.25 in / $3.75 out |
| 47 | Seed 2.1 Turbo seed-2.1-turbo multimodalvisionmulti-input reasoning | ByteDance | 59.1 overall | 64.1 | 0.0 | 58.5 | 53.4 | 0.0 | N/A |
| 48 | GPT-5.5 Pro gpt-5.5-pro multimodalvisionmulti-input reasoning | OpenAI | 59.1 overall | 60.1 | 0.0 | 67.5 | 48.8 | 0.0 | N/A |
| 49 | Gemini 4 Argon gemini-4-argon multimodalvisionmulti-input reasoning | Google | 59.0 overall | 0.0 | 85.2 | 0.0 | 0.0 | 17.0 | |
| 50 | Claude Opus 5.5 claude-opus-5-5 multimodalvisionmulti-input reasoning | Anthropic | 58.7 overall | 0.0 | 66.7 | 0.0 | 76.2 | 7.4 | |
| 51 | Gemini 3 Pro gemini-3-pro-preview multimodalvisionmulti-input reasoning | Google | 58.5 overall | 69.4 | 0.0 | 51.0 | 52.8 | 0.0 | |
| 52 | GLM-5.2 glm-5.2 codeprogrammingtool use | Zhipu AI | 58.4 overall | 67.2 | 85.2 | 39.9 | 55.9 | 40.7 | $0.75 in / $2.4 out |
| 53 | Gemini 3.8 Flash gemini-3.8-flash multimodalvisionmulti-input reasoning | Google | 58.4 overall | 52.3 | 85.2 | 0.0 | 0.0 | 32.7 | |
| 54 | Solar Mini 4 solar-mini4 codeprogrammingtool use | Upstage | 58.4 overall | 0.0 | 58.2 | 0.0 | 48.7 | 80.2 | $0.1 in / $0.4 out |
| 55 | Kimi K2-Thinking-0905 kimi-k2-thinking-0905 codeprogrammingtool use | Moonshot AI | 58.1 overall | 64.8 | 0.0 | 50.7 | 57.9 | 0.0 | |
| 56 | DeepSeek-V4-Pro-Max deepseek-v4-pro-max codeprogrammingtool use | DeepSeek | 58.0 overall | 64.3 | 85.2 | 46.9 | 52.2 | 35.4 | |
| 57 | Laguna S 2.1 laguna-s-2.1 codeprogrammingtool use | Poolside | 57.9 overall | 0.0 | 85.2 | 33.7 | 50.8 | 88.2 | $0.1 in / $0.2 out |
| 58 | MiniMax M3.1 Flash minimax-m3.1-flash multimodalvisionmulti-input reasoning | MiniMax | 57.5 overall | 47.5 | 66.7 | 0.0 | 0.0 | 70.9 | |
| 59 | DeepSeek-V3.2 (Non-thinking) deepseek-chat textinference | DeepSeek | 57.5 overall | 0.0 | 48.6 | 0.0 | 0.0 | 71.8 | $0.28 in / $0.42 out |
| 60 | GPT-5.1 Thinking gpt-5.1-thinking-2025-11-12 multimodalvisionmulti-input reasoning | OpenAI | 57.3 overall | 61.8 | 0.0 | 0.0 | 51.6 | 0.0 |
41
Ember-1
Fireworks AI
60.9
$3 in / $15 out
42
DeepSeek-V4.1-Flash-1017
DeepSeek
60.7
$0.3 in / $1.2 out
43
GPT-5.5
OpenAI
60.5
$5 in / $30 out
44