LiveCodeBench
Competitive-programming problem solving on problems published after a model's training cut-off, which limits contamination.
Code generationLiveCodeBench maintainers143 models run by Vals AIUpdated 1 September 2026
How it is run
Problems are pulled from live contest feeds and executed against hidden tests, so a solution has to compile and pass rather than merely look right.
What it cannot tell you
Single-function algorithmic tasks are a long way from repository maintenance, and the leaderboard is not deduplicated by date, so a 143-model board mixes many versions of the same family.
- Leader
- Claude Fable 5.190.52% · Vals AI's own list
- Top-ten spread
- 2.67%Leader minus tenth place, or last place in smaller fields.
- Median
- 87.35%of the 21 models on this site
- No published result
- 25of 46 models on this site
Independent leaderboard · Vals AI
Top 143 of 143 runs by Vals AI, one harness.
| # | Model | Score |
|---|---|---|
| 1 | Claude Fable 5.1 | 90.52% |
| 2 | Claude Fable 5 | 89.78% |
| 3 | Gemini 3.8 Flash | 89.48% |
| 4 | Claude Opus 5 | 89.03% |
| 5 | Gemini 3.7 Flash | 88.65% |
| 6 | Gemini 3.1 Pro | 88.49% |
| 7 | Grok 4.6 | 88.22% |
| 8 | Gemini 3.6 Flash | 88.08% |
| 9 | Gpt 5.2 Codex | 87.99% |
| 10 | Qwen3.8 Max | 87.85% |
| 11 | Claude Opus 4.8 | 87.82% |
| 12 | Gemini 3.5 Flash | 87.60% |
| 13 | DeepSeek V4 Pro | 87.53% |
| 14 | Deepseek V4 Pro | 87.48% |
| 15 | Grok 4.5 | 87.35% |
| 16 | Gpt 5.3 Codex | 87.31% |
| 17 | Deepseek V4 Flash 0731 | 87.26% |
| 18 | Kimi K3 | 87.19% |
| 19 | Qwen3.7 Max | 87.06% |
| 20 | Kimi K2.6 | 86.77% |
| 21 | Gpt 5 Mini 2025 08 07 | 86.61% |
| 22 | Gpt 5.1 2025 11 13 | 86.49% |
| 23 | Gemini 3 Pro Preview | 86.41% |
| 24 | Nemotron 3 Ultra 550b A55b | 85.98% |
| 25 | Qwen3.6 Plus | 85.95% |
| 26 | Inkling Small | 85.93% |
| 26 | GPT-5.6 Terra | 85.93% |
| 28 | Gpt 5 2025 08 07 | 85.91% |
| 28 | Muse Spark 1 1 | 85.91% |
| 30 | Gemini 3 Flash Preview | 85.59% |
| 31 | Gpt 5.1 Codex | 85.55% |
| 32 | Inkling | 85.52% |
| 33 | Gpt 5.2 2025 12 11 | 85.36% |
| 34 | Qwen3.5 Plus Thinking | 85.33% |
| 35 | GPT-5.5 | 85.30% |
| 36 | Claude Opus 4.7 | 85.07% |
| 37 | Gpt 5 Codex | 84.73% |
| 38 | Claude Opus 4 6 Thinking | 84.68% |
| 39 | Grok 4.3 | 84.49% |
| 40 | Grok 4.20 0309 Reasoning | 84.27% |
| 41 | Gpt 5.4 2026 03 05 | 84.14% |
| 42 | Gpt 5.4 Nano 2026 03 17 | 84.01% |
| 43 | Qwen3.8 27b | 84.00% |
| 44 | Ling 3.0 Flash 2607 | 83.99% |
| 45 | O3 2025 04 16 | 83.91% |
| 46 | Kimi K2.5 Thinking | 83.87% |
| 47 | Claude Opus 4 5 20251101 Thinking | 83.67% |
| 48 | Gpt 5.1 Codex Max | 83.56% |
| 49 | Qwen3.5 Flash | 83.28% |
| 50 | Grok 4 0709 | 83.25% |
| 51 | Gpt Oss 120b | 83.23% |
| 52 | GPT-5.6 Sol | 82.60% |
| 53 | Claude Sonnet 5 | 82.43% |
| 54 | Glm 4.7 | 82.23% |
| 55 | O4 Mini 2025 04 16 | 82.21% |
| 56 | MiniMax M3 | 82.15% |
| 57 | Claude Sonnet 4 6 | 82.09% |
| 58 | Kimi K2.7 Code | 82.05% |
| 59 | Glm 5 Thinking | 81.87% |
| 60 | MiniMax M2.1 | 81.76% |
| 61 | Mimo V2.5 | 81.51% |
| 62 | Gpt 5.4 Mini 2026 03 17 | 81.47% |
| 63 | Glm 5.1 | 81.38% |
| 64 | Mimo V2.5 Pro | 81.35% |
| 65 | Glm 4.6 | 81.04% |
| 66 | Deepseek V3p2 Thinking | 80.69% |
| 67 | Grok 4 1 Fast Reasoning | 80.64% |
| 68 | GLM 5.3 | 80.54% |
| 69 | GLM 5.3 Flash | 80.51% |
| 70 | Gpt Oss 20b | 80.39% |
| 71 | Gemini 3.1 Flash Lite Preview | 80.12% |
| 72 | MiniMax M2.7 | 79.93% |
| 73 | MiniMax M2.5 | 79.21% |
| 74 | Gemini 2.5 Pro Preview 03 25 | 79.16% |
| 75 | Gemini 3.5 Flash Lite | 79.01% |
| 76 | Grok 4 Fast Reasoning | 78.97% |
| 77 | Qwen3 Max | 78.22% |
| 78 | Grok 3 Mini Fast High Reasoning | 76.22% |
| 79 | Gemini 2.5 Flash Preview 09 2025 Thinking | 76.21% |
| 80 | Gemini 2.5 Flash Preview 09 2025 | 75.06% |
| 81 | Claude Opus 4 5 20251101 | 75.03% |
| 82 | Magistral Medium 2509 | 74.86% |
| 83 | Claude Sonnet 4 5 20250929 Thinking | 73.00% |
| 84 | Magistral Small 2509 | 72.13% |
| 85 | O3 Mini 2025 01 31 | 71.48% |
| 86 | Gemini 2.5 Flash Lite Preview 09 2025 Thinking | 71.39% |
| 87 | Qwen3 235b A22b | 70.62% |
| 88 | Kimi K2 Instruct | 70.45% |
| 89 | Deepseek R1 | 70.22% |
| 89 | Gpt 5 Nano 2025 08 07 | 70.22% |
| 91 | Claude Opus 4 20250514 Thinking | 70.19% |
| 92 | Deepseek V3p2 | 69.86% |
| 93 | Glm 5.2 | 69.50% |
| 94 | Laguna M.1 | 68.12% |
| 95 | Laguna Xs.2 | 67.80% |
| 96 | Gemini 2.5 Flash Lite Preview 09 2025 | 67.67% |
| 97 | Glm 4.5 | 67.45% |
| 98 | Qwen3 Max Preview | 66.91% |
| 99 | Claude Opus 4 1 20250805 Thinking | 66.46% |
| 100 | Grok 3 Mini Fast Low Reasoning | 66.27% |
| 101 | Deepseek V3 0324 | 65.48% |
| 102 | Claude Opus 4 1 20250805 | 64.56% |
| 103 | Kimi K2 Thinking | 63.15% |
| 104 | Claude Opus 4 20250514 | 62.63% |
| 105 | Claude Sonnet 4 20250514 Thinking | 62.39% |
| 106 | Grok Code Fast 1 | 61.97% |
| 107 | Claude 3 7 Sonnet 20250219 Thinking | 60.44% |
| 108 | Claude Sonnet 4 20250514 | 59.67% |
| 109 | Llama 3.3 Nemotron Super 49b V1 42e84561 Thinking | 58.37% |
| 110 | Gpt 4.1 Mini 2025 04 14 | 58.16% |
| 111 | Gemini 2.5 Flash Preview 04 17 | 56.94% |
| 112 | Claude 3 7 Sonnet 20250219 | 56.66% |
| 113 | Mistral Large 2512 | 55.34% |
| 114 | Gpt 4.1 2025 04 14 | 54.67% |
| 115 | Grok 3 | 52.90% |
| 116 | Devstral 2512 | 51.84% |
| 117 | O1 2024 12 17 | 50.26% |
| 118 | Claude 3 5 Sonnet 20241022 | 49.63% |
| 119 | Llama4 Maverick Instruct Basic | 47.25% |
| 120 | Gemini 2.5 Flash Preview 04 17 Thinking | 46.87% |
| 121 | Grok 4 Fast Non Reasoning | 46.10% |
| 122 | Mistral Medium 2505 | 44.85% |
| 123 | Gemini 2.0 Flash 001 | 43.61% |
| 124 | Gpt 4o 2024 11 20 | 43.44% |
| 125 | Labs Devstral Small 2512 | 43.18% |
| 126 | Gpt 4.1 Nano 2025 04 14 | 42.72% |
| 127 | Grok 4 1 Fast Non Reasoning | 42.62% |
| 128 | Claude 3 5 Haiku 20241022 | 41.92% |
| 129 | Gemini 1.5 Pro 002 | 41.72% |
| 130 | Claude Haiku 4 5 20251001 Thinking | 41.18% |
| 131 | Grok 2 1212 | 38.68% |
| 132 | Llama 4 Scout 17B 16E Instruct | 38.54% |
| 133 | Mistral Large 2411 | 37.09% |
| 134 | Gemini 1.5 Flash 002 | 36.91% |
| 135 | Llama 3.3 70B Instruct Turbo | 36.34% |
| 136 | Llama 3.3 Nemotron Super 49b V1 42e84561 | 36.31% |
| 137 | Command A 03 2025 | 35.07% |
| 138 | Mistral Small 2503 | 31.82% |
| 139 | Gpt 4o Mini 2024 07 18 | 26.42% |
| 140 | Jamba Large 1.6 | 22.33% |
| 141 | Command R Plus | 18.24% |
| 142 | Mistral Small 2402 | 15.78% |
| 143 | Jamba Mini 1.6 | 9.92% |
Artificial Analysis results by release
Highest published effort per model release, grouped here from AA’s public results. Grouping ranks are derived here; source settings and model-publisher reports stay separate.
| # | Model | Score | Effort |
|---|---|---|---|
| 1 | Gemini 3 Pro Preview | 91.75% | high |
| 2 | DeepSeek V3.2 Speciale | 89.63% | Not specified |
| 3 | GLM-4.7 | 89.42% | Not specified |
| 4 | GPT-5.2 | 88.89% | xhigh |
| 5 | gpt-oss-120b | 87.83% | high |
| 6 | Claude Opus 4.5 | 87.09% | Not specified |
| 7 | GPT-5.1 | 86.77% | high |
| 8 | o4-mini | 85.93% | high |
| 9 | Kimi K2 Thinking | 85.29% | Not specified |
| 10 | GPT-5.1 Codex | 84.87% | high |
| 11 | GPT-5 | 84.55% | high |
| 12 | GPT-5 Codex | 84.02% | high |
| 13 | GPT-5 mini | 83.81% | high |
| 14 | GPT-5.1 Codex mini | 83.60% | high |
| 15 | MiniMax-M2 | 82.65% | Not specified |
| 16 | Grok 4.1 Fast | 82.22% | Not specified |
| 17 | Grok 4 | 81.90% | Not specified |
| 18 | ERNIE 5.0 Thinking Preview | 81.16% | Not specified |
| 19 | MiniMax-M2.1 | 80.95% | Not specified |
| 20 | o3 | 80.85% | Not specified |
| 21 | Apriel-v1.6-15B-Thinker | 80.74% | Not specified |
| 22 | Gemini 2.5 Pro | 80.11% | Not specified |
| 23 | DeepSeek V3.1 Terminus | 79.79% | Not specified |
| 24 | Gemini 3 Flash Preview | 79.68% | Not specified |
| 25 | GPT-5 nano | 78.94% | high |
| 26 | Qwen3 235B A22B 2507 (Reasoning) | 78.84% | Not specified |
| 27 | DeepSeek V3.1 | 78.41% | Not specified |
| 27 | Qwen3 Next 80B A3B (Reasoning) | 78.41% | Not specified |
| 29 | Gemini 2.5 Pro Preview (Mar' 25) | 77.78% | Not specified |
| 30 | gpt-oss-20b | 77.67% | high |
| 30 | INTELLECT-3 | 77.67% | Not specified |
| 32 | Gemini 2.5 Pro Preview (May' 25) | 77.04% | Not specified |
| 32 | DeepSeek R1 0528 (May '25) | 77.04% | Not specified |
| 34 | K-EXAONE | 76.83% | Not specified |
| 35 | Qwen3 Max | 76.72% | Not specified |
| 36 | Doubao Seed Code | 76.61% | Not specified |
| 37 | Seed-OSS-36B-Instruct | 76.51% | Not specified |
| 38 | Magistral Medium 1.2 | 75.03% | Not specified |
| 39 | KAT-Coder-Pro V1 | 74.71% | Not specified |
| 40 | GLM-4.5 | 73.76% | Not specified |
| 40 | Qwen3 VL 32B (Reasoning) | 73.76% | Not specified |
| 42 | o3-mini | 73.44% | high |
| 43 | Nova 2.0 Pro Preview | 73.02% | medium |
| 44 | Apriel-v1.5-15B-Thinker | 72.80% | Not specified |
| 45 | Falcon-H1R-7B | 72.38% | Not specified |
| 46 | Magistral Small 1.2 | 72.28% | Not specified |
| 47 | Claude 4.5 Sonnet | 71.43% | Not specified |
| 48 | Nova 2.0 Lite | 71.11% | high |
| 48 | MiniMax M1 80k | 71.11% | Not specified |
| 50 | Qwen3 30B A3B 2507 (Reasoning) | 70.69% | Not specified |
| 51 | NVIDIA Nemotron Nano 9B V2 | 70.05% | Not specified |
| 52 | Qwen3 VL 30B A3B (Reasoning) | 69.74% | Not specified |
| 53 | Grok 3 mini Reasoning | 69.63% | high |
| 54 | Gemini 2.5 Flash | 69.52% | Not specified |
| 54 | Olmo 3.1 32B Think | 69.52% | Not specified |
| 54 | GLM-4.6 | 69.52% | Not specified |
| 57 | NVIDIA Nemotron Nano 12B v2 VL | 69.42% | Not specified |
| 57 | K2-V2 | 69.42% | high |
| 59 | Cogito v2.1 | 68.78% | Not specified |
| 60 | Hermes 4 - Llama-3.1 405B | 68.57% | Not specified |
| 61 | Qwen3 Next 80B A3B Instruct | 68.36% | Not specified |
| 61 | GLM-4.5-Air | 68.36% | Not specified |
| 63 | o1 | 67.94% | Not specified |
| 63 | Qwen3 Omni 30B A3B (Reasoning) | 67.94% | Not specified |
| 65 | Ling-1T | 67.72% | Not specified |
| 66 | Olmo 3 32B Think | 67.20% | Not specified |
| 67 | Nova 2.0 Omni | 66.03% | medium |
| 68 | MiniMax M1 40k | 65.71% | Not specified |
| 68 | Grok Code Fast 1 | 65.71% | Not specified |
| 70 | Mi:dm K 2.5 Pro | 65.61% | Not specified |
| 71 | Claude 4 Sonnet | 65.50% | Not specified |
| 72 | Claude 4.1 Opus | 65.35% | Not specified |
| 73 | Qwen3 Max (Preview) | 65.08% | Not specified |
| 73 | Motif-2-12.7B-Reasoning | 65.08% | Not specified |
| 75 | Qwen3 VL 235B A22B (Reasoning) | 64.55% | Not specified |
| 76 | Ring-1T | 64.34% | Not specified |
| 77 | Qwen3 4B 2507 (Reasoning) | 64.13% | Not specified |
| 77 | Gemini 2.5 Flash-Lite Preview (Sep '25) | 64.13% | Not specified |
| 77 | Llama 3.1 Nemotron Ultra 253B v1 | 64.13% | Not specified |
| 80 | QwQ 32B | 63.07% | Not specified |
| 81 | HyperCLOVA X SEED Think (32B) | 62.86% | Not specified |
| 82 | Ring-flash-2.0 | 62.75% | Not specified |
| 83 | Gemini 2.5 Flash Preview (Sep '25) | 62.54% | Not specified |
| 84 | DeepSeek R1 (Jan '25) | 61.69% | Not specified |
| 84 | Olmo 3 7B Think | 61.69% | Not specified |
| 86 | Solar Pro 2 | 61.59% | high |
| 87 | Claude 4.5 Haiku | 61.48% | Not specified |
| 88 | Kimi K2 0905 | 60.95% | Not specified |
| 89 | Qwen3 VL 235B A22B Instruct | 59.37% | Not specified |
| 90 | DeepSeek V3.2 | 59.26% | Not specified |
| 91 | Ling-flash-2.0 | 58.94% | Not specified |
| 92 | Qwen3 Coder 480B A35B Instruct | 58.52% | Not specified |
| 93 | o1-mini | 57.57% | Not specified |
| 93 | Mi:dm K 2.5 Pro Preview | 57.57% | Not specified |
| 95 | Kimi K2 | 55.56% | Not specified |
| 96 | DeepSeek V3.2 Exp | 55.45% | Not specified |
| 97 | Qwen3 32B | 54.60% | Not specified |
| 98 | GPT-5 (ChatGPT) | 54.29% | Not specified |
| 99 | Claude 4 Opus | 54.18% | Not specified |
| 100 | Qwen3 Max Thinking (Preview) | 53.54% | Not specified |
| 101 | Magistral Medium 1 | 52.70% | Not specified |
| 102 | Qwen3 235B A22B 2507 | 52.38% | Not specified |
| 103 | Qwen3 14B | 52.28% | Not specified |
| 104 | Qwen3 30B A3B 2507 | 51.53% | Not specified |
| 105 | Magistral Small 1 | 51.43% | Not specified |
| 105 | Qwen3 VL 32B Instruct | 51.43% | Not specified |
| 107 | DeepSeek R1 0528 Qwen3 8B | 51.32% | Not specified |
| 108 | Qwen3 30B A3B | 50.58% | Not specified |
| 109 | Llama 3.1 Nemotron Nano 4B v1.1 | 49.31% | Not specified |
| 110 | GPT-4.1 mini | 48.25% | Not specified |
| 111 | Qwen3 VL 30B A3B Instruct | 47.62% | Not specified |
| 112 | Claude 3.7 Sonnet | 47.30% | Not specified |
| 113 | EXAONE 4.0 32B | 47.20% | Not specified |
| 114 | ERNIE 4.5 300B A47B | 46.67% | Not specified |
| 115 | Qwen3 4B | 46.46% | Not specified |
| 115 | Mistral Large 3 | 46.46% | Not specified |
| 117 | Solar Pro 2 (Preview) | 46.24% | high |
| 118 | GPT-4.1 | 45.71% | Not specified |
| 119 | Devstral 2 | 44.76% | Not specified |
| 120 | Reka Flash 3 | 43.49% | Not specified |
| 121 | Ling-mini-2.0 | 42.86% | Not specified |
| 122 | Grok 3 | 42.54% | Not specified |
| 122 | GPT-4o (March 2025, chatgpt-4o-latest) | 42.54% | Not specified |
| 124 | Qwen3 Omni 30B A3B Instruct | 42.22% | Not specified |
| 125 | GLM-4.6V | 41.06% | Not specified |
| 126 | Mistral Medium 3.1 | 40.63% | Not specified |
| 126 | Gemini 2.5 Flash Preview | 40.63% | Not specified |
| 126 | Qwen3 8B | 40.63% | Not specified |
| 129 | DeepSeek V3 0324 | 40.53% | Not specified |
| 130 | Qwen3 Coder 30B A3B Instruct | 40.32% | Not specified |
| 131 | MiMo-V2-Flash | 40.21% | Not specified |
| 132 | Grok 4 Fast | 40.11% | Not specified |
| 133 | Gemini 2.5 Flash-Lite | 40.00% | Not specified |
| 133 | Mistral Medium 3 | 40.00% | Not specified |
| 135 | Llama 4 Maverick | 39.68% | Not specified |
| 136 | Claude 3.5 Sonnet (Oct '24) | 38.10% | Not specified |
| 137 | Kimi Linear 48B A3B Instruct | 37.78% | Not specified |
| 138 | Qwen3 4B 2507 | 37.67% | Not specified |
| 139 | DeepSeek R1 Distill Qwen 14B | 37.57% | Not specified |
| 140 | NVIDIA Nemotron 3 Nano 30B A3B | 35.98% | Not specified |
| 141 | DeepSeek V3 (Dec '24) | 35.87% | Not specified |
| 141 | Qwen2.5 Max | 35.87% | Not specified |
| 143 | Qwen3 VL 8B (Reasoning) | 35.34% | Not specified |
| 144 | GLM-4.5V | 35.24% | Not specified |
| 145 | Ministral 3 14B | 35.13% | Not specified |
| 146 | Devstral Small 2 | 34.81% | Not specified |
| 147 | Gemini 2.0 Pro Experimental (Feb '25) | 34.71% | Not specified |
| 148 | Qwen3 235B A22B | 34.29% | Not specified |
| 149 | QwQ 32B-Preview | 33.65% | Not specified |
| 149 | Devstral Medium | 33.65% | Not specified |
| 151 | Gemini 2.0 Flash (Feb '25) | 33.44% | Not specified |
| 151 | GPT-4o (May '24) | 33.44% | Not specified |
| 153 | Qwen3 VL 8B Instruct | 33.23% | Not specified |
| 154 | GPT-4.1 nano | 32.59% | Not specified |
| 155 | Gemini 2.0 Flash Thinking Experimental (Jan '25) | 32.06% | Not specified |
| 156 | Qwen3 VL 4B (Reasoning) | 31.96% | Not specified |
| 157 | GPT-4o (Aug '24) | 31.75% | Not specified |
| 157 | Nova Premier | 31.75% | Not specified |
| 159 | Gemini 1.5 Pro (Sep '24) | 31.64% | Not specified |
| 160 | Claude 3.5 Haiku | 31.43% | Not specified |
| 161 | GPT-4o (Nov '24) | 30.90% | Not specified |
| 162 | Llama 3.1 Instruct 405B | 30.48% | Not specified |
| 163 | Ministral 3 8B | 30.26% | Not specified |
| 164 | Llama 4 Scout | 29.95% | Not specified |
| 165 | Qwen2.5 Coder Instruct 32B | 29.52% | Not specified |
| 165 | Sarvam M | 29.52% | high |
| 165 | Sonar | 29.52% | Not specified |
| 168 | Mistral Large 2 (Nov '24) | 29.31% | Not specified |
| 168 | Exaone 4.0 1.2B | 29.31% | Not specified |
| 170 | GPT-4 Turbo | 29.10% | Not specified |
| 170 | Llama 3.1 Tulu3 405B | 29.10% | Not specified |
| 172 | Llama Nemotron Super 49B v1.5 | 28.99% | Not specified |
| 172 | Qwen3 VL 4B Instruct | 28.99% | Not specified |
| 174 | Llama 3.3 Instruct 70B | 28.78% | Not specified |
| 175 | Command A | 28.68% | Not specified |
| 176 | Claude 3 Opus | 27.94% | Not specified |
| 177 | Llama 3.3 Nemotron Super 49B v1 | 27.72% | Not specified |
| 178 | Qwen2.5 Instruct 72B | 27.62% | Not specified |
| 179 | Sonar Pro | 27.51% | Not specified |
| 179 | Mistral Small 3.2 | 27.51% | Not specified |
| 181 | Gemini 1.5 Flash (Sep '24) | 27.30% | Not specified |
| 182 | DeepSeek R1 Distill Qwen 32B | 26.98% | Not specified |
| 183 | Hermes 4 - Llama-3.1 70B | 26.88% | Not specified |
| 184 | Mistral Large 2 (Jul '24) | 26.67% | Not specified |
| 184 | Grok 2 (Dec '24) | 26.67% | Not specified |
| 186 | DeepSeek R1 Distill Llama 70B | 26.56% | Not specified |
| 186 | Olmo 3 7B Instruct | 26.56% | Not specified |
| 188 | Pixtral Large | 26.14% | Not specified |
| 189 | Devstral Small (May '25) | 25.82% | Not specified |
| 190 | Devstral Small (Jul '25) | 25.40% | Not specified |
| 191 | Mistral Small 3 | 25.19% | Not specified |
| 192 | Granite 4.0 H Small | 25.08% | Not specified |
| 193 | Qwen2.5 Instruct 32B | 24.76% | Not specified |
| 194 | Ministral 3 3B | 24.66% | Not specified |
| 195 | Gemini 1.5 Pro (May '24) | 24.44% | Not specified |
| 196 | Grok Beta | 24.13% | Not specified |
| 197 | GPT-4o mini | 23.39% | Not specified |
| 198 | DeepSeek R1 Distill Llama 8B | 23.28% | Not specified |
| 198 | Nova Pro | 23.28% | Not specified |
| 200 | Llama 3.1 Instruct 70B | 23.17% | Not specified |
| 201 | Phi-4 | 23.07% | Not specified |
| 202 | Gemini 1.5 Flash-8B | 21.69% | Not specified |
| 203 | Llama 3.2 Instruct 90B (Vision) | 21.38% | Not specified |
| 204 | Mistral Small 3.1 | 21.16% | Not specified |
| 205 | Jamba Reasoning 3B | 20.95% | Not specified |
| 205 | Gemini 2.0 Flash (experimental) | 20.95% | Not specified |
| 207 | Llama 3 Instruct 70B | 19.79% | Not specified |
| 208 | Gemini 1.5 Flash (May '24) | 19.58% | Not specified |
| 209 | Claude 2.1 | 19.47% | Not specified |
| 209 | DeepHermes 3 - Mistral 24B Preview | 19.47% | Not specified |
| 211 | Hermes 3 - Llama-3.1 70B | 18.84% | Not specified |
| 212 | Gemini 2.0 Flash-Lite (Feb '25) | 18.52% | Not specified |
| 213 | Jamba 1.7 Large | 18.10% | Not specified |
| 214 | Granite 4.0 Micro | 17.99% | Not specified |
| 215 | Gemini 2.0 Flash-Lite (Preview) | 17.88% | Not specified |
| 216 | Mistral Large (Feb '24) | 17.78% | Not specified |
| 217 | Claude 3 Sonnet | 17.46% | Not specified |
| 218 | Jamba 1.6 Large | 17.25% | Not specified |
| 219 | Claude 2.0 | 17.14% | Not specified |
| 220 | Llama 3.1 Nemotron Instruct 70B | 16.93% | Not specified |
| 221 | Nova Lite | 16.72% | Not specified |
| 222 | Qwen2.5 Turbo | 16.30% | Not specified |
| 223 | Qwen2 Instruct 72B | 15.87% | Not specified |
| 224 | DeepSeek Coder V2 Lite Instruct | 15.77% | Not specified |
| 225 | Claude 3 Haiku | 15.45% | Not specified |
| 226 | LFM2 8B A1B | 15.13% | Not specified |
| 227 | Mixtral 8x22B Instruct | 14.81% | Not specified |
| 228 | Gemma 3n E4B Instruct | 14.60% | Not specified |
| 229 | Jamba 1.5 Large | 14.29% | Not specified |
| 230 | Mistral Small (Sep '24) | 14.07% | Not specified |
| 231 | Nova Micro | 13.97% | Not specified |
| 232 | Gemma 3n E4B Instruct Preview (May '25) | 13.76% | Not specified |
| 233 | Gemma 3 27B Instruct | 13.65% | Not specified |
| 233 | Gemma 3 12B Instruct | 13.65% | Not specified |
| 235 | Phi-4 Multimodal Instruct | 13.12% | Not specified |
| 236 | Granite 3.3 8B | 12.70% | Not specified |
| 237 | Qwen3 1.7B | 12.59% | Not specified |
| 237 | Phi-4 Mini Instruct | 12.59% | Not specified |
| 237 | Qwen2.5 Coder Instruct 7B | 12.59% | Not specified |
| 240 | Command-R+ (Apr '24) | 12.17% | Not specified |
| 241 | Llama 3.1 Instruct 8B | 11.64% | Not specified |
| 241 | Gemini 1.0 Pro | 11.64% | Not specified |
| 241 | Phi-3 Mini Instruct 3.8B | 11.64% | Not specified |
| 244 | Granite 4.0 H 1B | 11.53% | Not specified |
| 244 | OpenChat 3.5 (1210) | 11.53% | Not specified |
| 246 | Gemma 3 4B Instruct | 11.22% | Not specified |
| 247 | Mistral Small (Feb '24) | 11.11% | Not specified |
| 248 | Llama 3.2 Instruct 11B (Vision) | 11.01% | Not specified |
| 249 | Claude Instant | 10.90% | Not specified |
| 250 | Mistral Medium | 9.95% | Not specified |
| 251 | Llama 2 Chat 70B | 9.84% | Not specified |
| 251 | Llama 2 Chat 13B | 9.84% | Not specified |
| 253 | LFM 40B | 9.63% | Not specified |
| 253 | Llama 3 Instruct 8B | 9.63% | Not specified |
| 255 | Gemma 3n E2B Instruct | 9.52% | Not specified |
| 256 | DBRX Instruct | 9.31% | Not specified |
| 257 | DeepHermes 3 - Llama-3.1 8B Preview | 8.47% | Not specified |
| 258 | Llama 3.2 Instruct 3B | 8.25% | Not specified |
| 259 | LFM2 2.6B | 8.15% | Not specified |
| 260 | Qwen3 0.6B | 7.30% | Not specified |
| 261 | Jamba 1.6 Mini | 7.09% | Not specified |
| 262 | DeepSeek R1 Distill Qwen 1.5B | 6.98% | Not specified |
| 263 | OLMo 2 32B | 6.77% | Not specified |
| 264 | Mixtral 8x7B Instruct | 6.56% | Not specified |
| 265 | Jamba 1.5 Mini | 6.24% | Not specified |
| 266 | Jamba 1.7 Mini | 6.14% | Not specified |
| 267 | Command-R (Mar '24) | 4.76% | Not specified |
| 268 | Granite 4.0 1B | 4.66% | Not specified |
| 269 | Mistral 7B Instruct | 4.55% | Not specified |
| 270 | OLMo 2 7B | 4.13% | Not specified |
| 271 | Molmo 7B-D | 3.92% | Not specified |
| 272 | Granite 4.0 350M | 2.43% | Not specified |
| 273 | LFM2 1.2B | 2.01% | Not specified |
| 274 | Llama 3.2 Instruct 1B | 1.90% | Not specified |
| 274 | Granite 4.0 H 350M | 1.90% | Not specified |
| 276 | Gemma 3 1B Instruct | 1.69% | Not specified |
| 277 | Gemma 3 270M | 0.32% | Not specified |
| 278 | Llama 2 Chat 7B | 0.21% | Not specified |
No published result tracked
25 of the 46 models on this site have no published number on this benchmark, so their cells in the home table stay blank rather than reading as zero.
Step 5 PreviewGPT-6 LunaQwen3.8 2.4T A95BQwen3.8-Flash-NextMiMo-V2.6-FlashQwen3.8 Max (0902)Muse Spark 1.2GPT-5.4DeepSeek V4.1 Flash 1017GLM-LegalClaude Haiku 5.5BeamGemini 4 ArgonClaude Sonnet 5.5Claude Opus 5.5GPT-6 AstraGPT-6 SolGPT-6.1 SolGPT-5.6 LunaGrok 4.7DeepSeek V4.1 FlashMuse Spark 1.3 MaxHunyuan Hy4 PreviewMiMo V2.6 ProMistral Medium 3.5
Sources
Every page used for this benchmark, read on 17 October 2026.
- Benchmark site or the page that publishes itlivecodebench.github.io/
- Independent resultswww.vals.ai/benchmarks/lcb