Competitive coding and code-generation benchmark score.
LiveCodeBench collects recent problems from platforms such as LeetCode, AtCoder, and Codeforces. Its release-date tracking helps evaluate models on newer, less contaminated coding tasks.
Test type: Code generation and related coding scenarios, scored with executable tests.
343 models have this metric.
Current leader: Gemini 3 Pro Preview (high)
Project links
This app ranks the LiveCodeBench score exposed by the Artificial Analysis snapshot.
Top models ranked by LCB.
| Rank | Model | Creator | Value | Speed | Blended Price |
|---|---|---|---|---|---|
| #1 | Gemini 3 Pro Preview (high) | 91.7% | 128.7 tok/s | $4.50/M | |
| #2 |
| 90.8% |
| 193.2 tok/s |
| $1.13/M |
| #3 | DeepSeek V3.2 Speciale | DeepSeek | 89.6% | n/a | - |
| #4 | GLM-4.7 (Reasoning) | Z AI | 89.4% | 90.3 tok/s | $1.00/M |
| #5 | GPT-5.2 (medium) | OpenAI | 89.4% | n/a | $4.81/M |
| #6 | GPT-5.2 (xhigh) | OpenAI | 88.9% | 71.8 tok/s | $4.81/M |
| #7 | gpt-oss-120B (high) | OpenAI | 87.8% | 212.3 tok/s | $0.263/M |
| #8 | Claude Opus 4.5 (Reasoning) | Anthropic | 87.1% | 57 tok/s | $10.00/M |
| #9 | GPT-5.1 (high) | OpenAI | 86.8% | 123.3 tok/s | $3.44/M |
| #10 | MiMo-V2-Flash (Reasoning) | Xiaomi | 86.8% | 118.8 tok/s | $0.150/M |
| #11 | DeepSeek V3.2 (Reasoning) | DeepSeek | 86.2% | n/a | $0.315/M |
| #12 | o4-mini (high) | OpenAI | 85.9% | 124.5 tok/s | $1.93/M |
| #13 | Gemini 3 Pro Preview (low) | 85.7% | n/a | $4.50/M |
| #14 | Kimi K2 Thinking | Kimi | 85.3% | 99 tok/s | $1.08/M |
| #15 | GPT-5.1 Codex (high) | OpenAI | 84.9% | 162.7 tok/s | $3.44/M |
| #16 | GPT-5 (high) | OpenAI | 84.6% | 84.2 tok/s | $3.44/M |
| #17 | GPT-5 Codex (high) | OpenAI | 84.0% | 166.8 tok/s | $3.44/M |
| #18 | GPT-5 mini (high) | OpenAI | 83.8% | 85.7 tok/s | $0.688/M |
| #19 | GPT-5.1 Codex mini (high) | OpenAI | 83.6% | 207.2 tok/s | $0.688/M |
| #20 | Grok 4 Fast (Reasoning) | xAI | 83.2% | 76.2 tok/s | $0.275/M |
| #21 | MiniMax-M2 | MiniMax | 82.6% | 83.5 tok/s | $0.525/M |
| #22 | Grok 4.1 Fast (Reasoning) | xAI | 82.2% | 140.9 tok/s | $0.275/M |
| #23 | Grok 4 | xAI | 81.9% | 50.3 tok/s | $6.00/M |
| #24 | ERNIE 5.0 Thinking Preview | Baidu | 81.2% | n/a | - |
| #25 | MiniMax-M2.1 | MiniMax | 81.0% | 84.8 tok/s | $0.525/M |
| #26 | o3 | OpenAI | 80.8% | 72.7 tok/s | $3.50/M |
| #27 | Apriel-v1.6-15B-Thinker | ServiceNow | 80.7% | n/a | - |
| #28 | Gemini 2.5 Pro | 80.1% | 120.2 tok/s | $3.44/M |
| #29 | DeepSeek V3.1 Terminus (Reasoning) | DeepSeek | 79.8% | n/a | $1.91/M |
| #30 | Gemini 3 Flash Preview (Non-reasoning) | 79.7% | 178.3 tok/s | $1.13/M |
| #31 | DeepSeek V3.2 Exp (Reasoning) | DeepSeek | 78.9% | n/a | $0.315/M |
| #32 | GPT-5 nano (high) | OpenAI | 78.9% | 136 tok/s | $0.138/M |
| #33 | Qwen3 235B A22B 2507 (Reasoning) | Alibaba | 78.8% | 56 tok/s | $2.63/M |
| #34 | DeepSeek V3.1 (Reasoning) | DeepSeek | 78.4% | n/a | $0.865/M |
| #35 | Qwen3 Next 80B A3B (Reasoning) | Alibaba | 78.4% | 172.2 tok/s | $1.88/M |
| #36 | Gemini 2.5 Pro Preview (Mar' 25) | 77.8% | n/a | - |
| #37 | gpt-oss-20B (high) | OpenAI | 77.7% | 242.3 tok/s | $0.100/M |
| #38 | INTELLECT-3 | Prime Intellect | 77.7% | n/a | - |
| #39 | DeepSeek R1 0528 (May '25) | DeepSeek | 77.0% | n/a | $2.36/M |
| #40 | Gemini 2.5 Pro Preview (May' 25) | 77.0% | n/a | $3.44/M |
| #41 | K-EXAONE (Reasoning) | LG AI Research | 76.8% | n/a | - |
| #42 | Qwen3 Max | Alibaba | 76.7% | 32.2 tok/s | $2.40/M |
| #43 | Doubao Seed Code | ByteDance Seed | 76.6% | n/a | - |
| #44 | Seed-OSS-36B-Instruct | ByteDance Seed | 76.5% | 40 tok/s | $0.300/M |
| #45 | GPT-5 (low) | OpenAI | 76.3% | 65.8 tok/s | $3.44/M |
| #46 | GPT-5 nano (medium) | OpenAI | 76.3% | 150.3 tok/s | $0.138/M |
| #47 | Magistral Medium 1.2 | Mistral | 75.0% | 42 tok/s | $2.75/M |
| #48 | EXAONE 4.0 32B (Reasoning) | LG AI Research | 74.7% | n/a | - |
| #49 | KAT-Coder-Pro V1 | KwaiKAT | 74.7% | 117.1 tok/s | $0.525/M |
| #50 | NVIDIA Nemotron 3 Nano 30B A3B (Reasoning) | NVIDIA | 74.1% | 154.8 tok/s | $0.096/M |
| #51 | Claude Opus 4.5 (Non-reasoning) | Anthropic | 73.8% | 50.3 tok/s | $10.00/M |
| #52 | GLM-4.5 (Reasoning) | Z AI | 73.8% | 46.4 tok/s | $1.00/M |
| #53 | Qwen3 VL 32B (Reasoning) | Alibaba | 73.8% | 94.5 tok/s | $2.63/M |
| #54 | Llama Nemotron Super 49B v1.5 (Reasoning) | NVIDIA | 73.7% | 50.8 tok/s | $0.175/M |
| #55 | o3-mini (high) | OpenAI | 73.4% | 140 tok/s | $1.93/M |
| #56 | Nova 2.0 Pro Preview (medium) | Amazon | 73.0% | 112.7 tok/s | $3.44/M |
| #57 | Apriel-v1.5-15B-Thinker | ServiceNow | 72.8% | n/a | - |
| #58 | Falcon-H1R-7B | TII UAE | 72.4% | n/a | - |
| #59 | NVIDIA Nemotron Nano 9B V2 (Reasoning) | NVIDIA | 72.4% | 121.6 tok/s | $0.070/M |
| #60 | Magistral Small 1.2 | Mistral | 72.3% | 100.3 tok/s | $0.750/M |
| #61 | o3-mini | OpenAI | 71.7% | 140.1 tok/s | $1.93/M |
| #62 | Claude 4.5 Sonnet (Reasoning) | Anthropic | 71.4% | 43.8 tok/s | $6.00/M |
| #63 | Gemini 2.5 Flash Preview (Sep '25) (Reasoning) | 71.3% | n/a | - |
| #64 | MiniMax M1 80k | MiniMax | 71.1% | n/a | $0.963/M |
| #65 | Nova 2.0 Lite (high) | Amazon | 71.1% | 170.7 tok/s | $0.850/M |
| #66 | gpt-oss-120B (low) | OpenAI | 70.7% | 216.3 tok/s | $0.263/M |
| #67 | Qwen3 30B A3B 2507 (Reasoning) | Alibaba | 70.7% | 143.2 tok/s | $0.750/M |
| #68 | GPT-5 (medium) | OpenAI | 70.3% | 82.3 tok/s | $3.44/M |
| #69 | NVIDIA Nemotron Nano 9B V2 (Non-reasoning) | NVIDIA | 70.1% | 153.3 tok/s | $0.086/M |
| #70 | Qwen3 VL 30B A3B (Reasoning) | Alibaba | 69.7% | 121.9 tok/s | $0.750/M |
| #71 | Grok 3 mini Reasoning (high) | xAI | 69.6% | 215.5 tok/s | $0.350/M |
| #72 | Gemini 2.5 Flash (Reasoning) | 69.5% | 199.6 tok/s | $0.850/M |
| #73 | GLM-4.6 (Reasoning) | Z AI | 69.5% | 26.3 tok/s | $0.963/M |
| #74 | Olmo 3.1 32B Think | Allen Institute for AI | 69.5% | n/a | - |
| #75 | K2-V2 (high) | MBZUAI Institute of Foundation Models | 69.4% | n/a | - |
| #76 | NVIDIA Nemotron Nano 12B v2 VL (Reasoning) | NVIDIA | 69.4% | 125 tok/s | $0.300/M |
| #77 | GPT-5 mini (medium) | OpenAI | 69.2% | 77.2 tok/s | $0.688/M |
| #78 | Cogito v2.1 (Reasoning) | Deep Cogito | 68.8% | 51.1 tok/s | $1.25/M |
| #79 | Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning) | 68.8% | n/a | $0.175/M |
| #80 | Hermes 4 - Llama-3.1 405B (Reasoning) | Nous Research | 68.6% | 34.9 tok/s | $1.50/M |