51 o1 aider-reasoning-high, high · Aider LLM Leaderboards 61.7% 52 o1 high · Epoch AI Benchmarking Hub data 61.7% 53 Claude Sonnet 4 32k · Epoch AI Benchmarking Hub data 61.3% 54 Claude Sonnet 4 aider-reasoning-32k, 32k thinking · Aider LLM Leaderboards 61.3% 55 Claude Sonnet 3.7 · Epoch AI Benchmarking Hub data 60.4% 56 Claude Sonnet 3.7 aider-reasoning-off, no thinking · Aider LLM Leaderboards 60.4% 57 o3 Mini aider-reasoning-high, high · Aider LLM Leaderboards 60.4% 58 o3 Mini high · Epoch AI Benchmarking Hub data 60.4% 59 Kimi K2 Instruct · LLM Stats data · self-reported Reported by the model's developer. The source relaying it did not verify it. claim Values from 2 sources 60% 60 Qwen 3 235B A22B · Epoch AI Benchmarking Hub data 59.6% 61 Qwen3 235B A22B Instruct 2507 · Epoch AI Benchmarking Hub data Values from 2 sources 59.6% 62 Kimi K2 aider · Aider LLM Leaderboards 59.1% 63 Kimi K2 0905 · Epoch AI Benchmarking Hub data 59.1% 64 DeepSeek R1 aider · Aider LLM Leaderboards 56.9% 65 Claude Sonnet 4 · Epoch AI Benchmarking Hub data 56.4% 66 Claude Sonnet 4 aider-reasoning-off, no thinking · Aider LLM Leaderboards 56.4% 67 DeepSeek V3 · Epoch AI Benchmarking Hub data Values from 2 sources 55.1% 68 DeepSeek V3 aider · Aider LLM Leaderboards 55.1% 69 Gemini 2.5 Flash Preview 05-20 (no think) 23k · Epoch AI Benchmarking Hub data 55.1% 70 Gemini 2.5 Flash Preview 05-20 (no think) aider-reasoning-24k, 24k think · Aider LLM Leaderboards 55.1% 71 o3 Mini aider-reasoning-medium, medium · Aider LLM Leaderboards 53.8% 72 o3 Mini medium · Epoch AI Benchmarking Hub data 53.8% 73 Grok 3 aider · Aider LLM Leaderboards 53.3% 74 Grok 3 beta · Epoch AI Benchmarking Hub data 53.3% 75 GPT 4.1 · Epoch AI Benchmarking Hub data Values from 2 sources 52.4% 76 GPT 4.1 aider · Aider LLM Leaderboards 52.4% 77 Claude 3.5 Sonnet · Epoch AI Benchmarking Hub data 51.6% 78 Claude 3.5 Sonnet aider · Aider LLM Leaderboards 51.6% 79 Qwen3 Next 80B A3B Instruct · LLM Stats data · self-reported Reported by the model's developer. The source relaying it did not verify it. claim 49.8% 80 Grok 3 Mini aider-reasoning-high, high · Aider LLM Leaderboards 49.3% 81 Grok 3 Mini high, beta · Epoch AI Benchmarking Hub data 49.3% 82 DeepSeek V3 aider, prev · Aider LLM Leaderboards 48.4% 83 gemini-2.5-flash-preview-04-17 · Epoch AI Benchmarking Hub data 47.1% 84 gemini-2.5-flash-preview-04-17 aider, default · Aider LLM Leaderboards 47.1% 85 Magistral Medium · LLM Stats data · self-reported Reported by the model's developer. The source relaying it did not verify it. claim 47.1% 86 GPT-4.5 Preview · Epoch AI Benchmarking Hub data 44.9% 87 GPT-4.5 Preview aider · Aider LLM Leaderboards 44.9% 88 GPT OSS 120B official model card Table 3 · reasoning high · published setting; Table 3 does not further specify tool access · pass@1 · Reported by OpenAI · via BenchBoard docs · self-reported Reported by the model's developer. The source relaying it did not verify it. 44.4% 89 Gemini 2.5 Flash Preview 05-20 (no think) · Epoch AI Benchmarking Hub data 44% 90 Gemini 2.5 Flash Preview 05-20 (no think) aider-reasoning-off, no think · Aider LLM Leaderboards 44% 91 GPT OSS 120B aider-reasoning-high, high · Aider LLM Leaderboards 41.8% 92 GPT OSS 120B high · Epoch AI Benchmarking Hub data 41.8% 93 Qwen3 32B · Epoch AI Benchmarking Hub data 40% 94 Qwen3 32B aider · Aider LLM Leaderboards 40% 95 gemini-exp · Epoch AI Benchmarking Hub data 38.2% 96 gemini-exp aider · Aider LLM Leaderboards 38.2% 97 gemini-2.0-pro-exp-02-05 · Epoch AI Benchmarking Hub data 35.6% 98 gemini-2.0-pro-exp-02-05 aider · Aider LLM Leaderboards 35.6% 99 GPT-4.1 mini · LLM Stats data · self-reported Reported by the model's developer. The source relaying it did not verify it. claim Values from 2 sources 34.7% 100 Grok 3 Mini aider-reasoning-low, low · Aider LLM Leaderboards 34.7%