Models with measured (third-party leaderboard) or author-reported evidence for "Reason through hard problems", dated by when the evidence was observed.https://openmodeldrops.com/enLiquidAI/LFM2.5-230M has author-reported evidence for "Reason through hard problems": Author reports mmlu_pro 20.25. Source: https://huggingface.co/LiquidAI/LFM2.5-230M/raw/main/.eval_results/LFM2.5-230M.yaml. Observed at 2026-10-24T21:42:38Z.Sat, 24 Oct 2026 21:42:38 GMThttps://openmodeldrops.com/models/LiquidAI/LFM2.5-230M/7d622a04f57507b26587LiquidAI/LFM2.5-350M has author-reported evidence for "Reason through hard problems": Author reports mmlu_pro 20.01. Source: https://huggingface.co/LiquidAI/LFM2.5-350M/raw/main/.eval_results/mmlu_pro.yaml. Observed at 2026-10-24T21:39:39Z.Sat, 24 Oct 2026 21:39:39 GMThttps://openmodeldrops.com/models/LiquidAI/LFM2.5-350M/51a5ab204028291c0176thinkingmachines/Inkling-Small has author-reported evidence for "Reason through hard problems": Author reports MathArena/aime_2026 95.5. Source: https://huggingface.co/thinkingmachines/Inkling-Small/raw/main/.eval_results/aime_2026.yaml. Observed at 2026-10-24T21:39:08Z.Sat, 24 Oct 2026 21:39:08 GMThttps://openmodeldrops.com/models/thinkingmachines/Inkling-Small/8c478df8108dbc714292datalab-to/surya-ocr-2 has author-reported evidence for "Reason through hard problems": Author reports arxiv_math 88.3. Source: https://huggingface.co/datalab-to/surya-ocr-2/raw/main/.eval_results/olmocrbench.yaml. Observed at 2026-10-24T21:38:52Z.Sat, 24 Oct 2026 21:38:52 GMThttps://openmodeldrops.com/models/datalab-to/surya-ocr-2/893994f31522767c97b6inclusionAI/Ling-3.0-flash has author-reported evidence for "Reason through hard problems": Author reports MathArena/aime_2026 93.2. Source: https://huggingface.co/inclusionAI/Ling-3.0-flash/raw/main/.eval_results/aime_2026.yaml. Observed at 2026-10-24T21:38:51Z.Sat, 24 Oct 2026 21:38:51 GMThttps://openmodeldrops.com/models/inclusionAI/Ling-3.0-flash/46ebdca5b0c2d6c41a90lightonai/LightOnOCR-2-1B has author-reported evidence for "Reason through hard problems": Author reports arxiv_math 89.6. Source: https://huggingface.co/lightonai/LightOnOCR-2-1B/raw/main/.eval_results/olmocrbench.yaml. Observed at 2026-10-24T21:38:42Z.Sat, 24 Oct 2026 21:38:42 GMThttps://openmodeldrops.com/models/lightonai/LightOnOCR-2-1B/8c8b376babde68094d25LiquidAI/LFM2.5-8B-A1B has author-reported evidence for "Reason through hard problems": Author reports MathArena/aime_2026 50.0. Source: https://huggingface.co/LiquidAI/LFM2.5-8B-A1B/raw/main/.eval_results/aime_2026.yaml. Observed at 2026-10-24T21:38:26Z.Sat, 24 Oct 2026 21:38:26 GMThttps://openmodeldrops.com/models/LiquidAI/LFM2.5-8B-A1B/daf0ffd2440b9734f01bdatalab-to/chandra-ocr-2 has author-reported evidence for "Reason through hard problems": Author reports old_scans_math 89.1. Source: https://huggingface.co/datalab-to/chandra-ocr-2/raw/main/.eval_results/olmocrbench.yaml. Observed at 2026-10-24T21:37:56Z.Sat, 24 Oct 2026 21:37:56 GMThttps://openmodeldrops.com/models/datalab-to/chandra-ocr-2/1f64fd1186fb6926288bbaidu/Qianfan-OCR has author-reported evidence for "Reason through hard problems": Author reports arxiv_math 80.1. Source: https://huggingface.co/baidu/Qianfan-OCR/raw/main/.eval_results/olm_ocr_bench.yaml. Observed at 2026-10-24T21:37:34Z.Sat, 24 Oct 2026 21:37:34 GMThttps://openmodeldrops.com/models/baidu/Qianfan-OCR/59d327c468b3d5f2f427MiniMaxAI/MiniMax-M2.5 has author-reported evidence for "Reason through hard problems": Author reports hle 19.4. Source: https://huggingface.co/MiniMaxAI/MiniMax-M2.5/raw/main/.eval_results/hle.yaml. Observed at 2026-10-24T21:33:17Z.Sat, 24 Oct 2026 21:33:17 GMThttps://openmodeldrops.com/models/MiniMaxAI/MiniMax-M2.5/014f7731297a8d649ee7moonshotai/Kimi-K2.6 has author-reported evidence for "Reason through hard problems": Author reports MathArena/aime_2026 96.4. Source: https://huggingface.co/moonshotai/Kimi-K2.6/raw/main/.eval_results/aime_2026.yaml. Observed at 2026-10-24T21:32:01Z.Sat, 24 Oct 2026 21:32:01 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-K2.6/5a179c515f07716418d3zai-org/GLM-5.3 has author-reported evidence for "Reason through hard problems": Author reports hle 62.5 (With tools.). Source: https://huggingface.co/zai-org/GLM-5.3/raw/main/.eval_results/hle.yaml. Observed at 2026-10-24T21:31:29Z.Sat, 24 Oct 2026 21:31:29 GMThttps://openmodeldrops.com/models/zai-org/GLM-5.3/ecd89cd5a4d3395d4b73zai-org/GLM-5.3-Flash has author-reported evidence for "Reason through hard problems": Author reports hle 55.3 (HLE with tools (full set) and a 300K-context management strategy, not the no-tools default; judged by GPT-5.6-luna (medium).). Source: https://huggingface.co/zai-org/GLM-5.3-Flash/raw/main/.eval_results/GLM-5.3-Flash.yaml. Observed at 2026-10-24T21:31:22Z.Sat, 24 Oct 2026 21:31:22 GMThttps://openmodeldrops.com/models/zai-org/GLM-5.3-Flash/3dc192797534997281e1perplexity-ai/pplx-decider-v1-27b has author-reported evidence for "Reason through hard problems": Author reports BBH 82.80%. Source: https://huggingface.co/perplexity-ai/pplx-decider-v1-27b/raw/main/README.md. Observed at 2026-10-14T14:06:59.654976+00:00.Wed, 14 Oct 2026 14:06:59 GMThttps://openmodeldrops.com/models/perplexity-ai/pplx-decider-v1-27b/295613386242e8aa4943kirp/jpt-9b has author-reported evidence for "Reason through hard problems": Author reports BBH 55.6 (Knowledge). Source: https://huggingface.co/kirp/jpt-9b/raw/main/README.md. Observed at 2026-10-07T20:26:22.147729+00:00.Wed, 07 Oct 2026 20:26:22 GMThttps://openmodeldrops.com/models/kirp/jpt-9b/a25e3f745e6c4011b901kirp/jpt-4b has author-reported evidence for "Reason through hard problems": Author reports Knowledge GPQA Diamond 23.8. Source: https://huggingface.co/kirp/jpt-4b/raw/main/README.md. Observed at 2026-10-07T20:25:54.610774+00:00.Wed, 07 Oct 2026 20:25:54 GMThttps://openmodeldrops.com/models/kirp/jpt-4b/9be9695b9da7d89ea858jaredpalmer/kev-0.8b has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro, 10 options (transfer-v9 development) 0.230. Source: https://huggingface.co/jaredpalmer/kev-0.8b/raw/main/README.md. Observed at 2026-10-07T20:25:33.848011+00:00.Wed, 07 Oct 2026 20:25:33 GMThttps://openmodeldrops.com/models/jaredpalmer/kev-0.8b/ea1072b9b48f131b110ajaredpalmer/kev-9b has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro, 10 options (transfer-v9 development) 0.590. Source: https://huggingface.co/jaredpalmer/kev-9b/raw/main/README.md. Observed at 2026-10-07T20:25:30.327089+00:00.Wed, 07 Oct 2026 20:25:30 GMThttps://openmodeldrops.com/models/jaredpalmer/kev-9b/f3d465655aad1562ea32jaredpalmer/kev-4b has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro, 10 options (transfer-v9 development) 0.565. Source: https://huggingface.co/jaredpalmer/kev-4b/raw/main/README.md. Observed at 2026-10-07T20:25:18.706101+00:00.Wed, 07 Oct 2026 20:25:18 GMThttps://openmodeldrops.com/models/jaredpalmer/kev-4b/247faef1b4db0e3fb2c5Qwen/Qwen2-72B has author-reported evidence for "Reason through hard problems": Author reports MMLU 84.2. Source: https://huggingface.co/Qwen/Qwen2-72B/raw/main/README.md. Observed at 2026-10-07T20:24:34.334145+00:00.Wed, 07 Oct 2026 20:24:34 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-72B/65c100b5c934b4e0b855RadixArk/Kimi-K3-DSpark has author-reported evidence for "Reason through hard problems": Author reports GSM8K 5.4176. Source: https://huggingface.co/RadixArk/Kimi-K3-DSpark/raw/main/README.md. Observed at 2026-10-07T20:24:31.229018+00:00.Wed, 07 Oct 2026 20:24:31 GMThttps://openmodeldrops.com/models/RadixArk/Kimi-K3-DSpark/f285857a7428b9c7ca83internlm/internlm3-8b-instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU(0-shot) 76.6. Source: https://huggingface.co/internlm/internlm3-8b-instruct/raw/main/README.md. Observed at 2026-10-07T20:24:30.589966+00:00.Wed, 07 Oct 2026 20:24:30 GMThttps://openmodeldrops.com/models/internlm/internlm3-8b-instruct/1a82a91723ecbce5fd97microsoft/Phi-tiny-MoE-instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 60.83 (Phi-tiny-MoE). Source: https://huggingface.co/microsoft/Phi-tiny-MoE-instruct/raw/main/README.md. Observed at 2026-10-07T20:24:27.695265+00:00.Wed, 07 Oct 2026 20:24:27 GMThttps://openmodeldrops.com/models/microsoft/Phi-tiny-MoE-instruct/841688d6b5a29395c42cupstage/Solar-Open-100B has author-reported evidence for "Reason through hard problems": Author reports KMMLU 73.0. Source: https://huggingface.co/upstage/Solar-Open-100B/raw/main/README.md. Observed at 2026-10-07T20:24:27.470529+00:00.Wed, 07 Oct 2026 20:24:27 GMThttps://openmodeldrops.com/models/upstage/Solar-Open-100B/4de4aaa071961edcbdbbibm-granite/granite-4.0-h-small has author-reported evidence for "Reason through hard problems": Author reports MMLU 78.44 (5-shot). Source: https://huggingface.co/ibm-granite/granite-4.0-h-small/raw/main/README.md. Observed at 2026-10-07T20:24:19.112612+00:00.Wed, 07 Oct 2026 20:24:19 GMThttps://openmodeldrops.com/models/ibm-granite/granite-4.0-h-small/70f6842e391b511755c5LGAI-EXAONE/K-EXAONE-236B-A23B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 83.8 (Reasoning). Source: https://huggingface.co/LGAI-EXAONE/K-EXAONE-236B-A23B/raw/main/README.md. Observed at 2026-10-07T20:24:16.249823+00:00.Wed, 07 Oct 2026 20:24:16 GMThttps://openmodeldrops.com/models/LGAI-EXAONE/K-EXAONE-236B-A23B/91c2731c8261d54dea82allenai/Olmo-3-7B-Think has author-reported evidence for "Reason through hard problems": Author reports MATH 95.1 (Olmo 3 Think 7B). Source: https://huggingface.co/allenai/Olmo-3-7B-Think/raw/main/README.md. Observed at 2026-10-07T20:24:14.562201+00:00.Wed, 07 Oct 2026 20:24:14 GMThttps://openmodeldrops.com/models/allenai/Olmo-3-7B-Think/b9102d3a9deda8110947Nanbeige/Nanbeige4.1-3B has author-reported evidence for "Reason through hard problems": Author reports AIME 2026 I 87.40. Source: https://huggingface.co/Nanbeige/Nanbeige4.1-3B/raw/main/README.md. Observed at 2026-10-07T20:24:14.555408+00:00.Wed, 07 Oct 2026 20:24:14 GMThttps://openmodeldrops.com/models/Nanbeige/Nanbeige4.1-3B/f79615dad37c1ce76bc6nvidia/Llama-3_3-Nemotron-Super-49B-v1 has author-reported evidence for "Reason through hard problems": Author reports MATH500 96.6 (Reasoning On). Source: https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1/raw/main/README.md. Observed at 2026-10-07T20:24:13.624816+00:00.Wed, 07 Oct 2026 20:24:13 GMThttps://openmodeldrops.com/models/nvidia/Llama-3_3-Nemotron-Super-49B-v1/97535141a265d298b6ebtencent/Hy3-preview has author-reported evidence for "Reason through hard problems": Author reports MMLU 87.42 (5-shot). Source: https://huggingface.co/tencent/Hy3-preview/raw/main/README.md. Observed at 2026-10-07T20:24:08.909803+00:00.Wed, 07 Oct 2026 20:24:08 GMThttps://openmodeldrops.com/models/tencent/Hy3-preview/80ccb8a74392e2f3c59dibm-granite/granite-4.0-h-tiny has author-reported evidence for "Reason through hard problems": Author reports MMLU 68.65 (5-shot). Source: https://huggingface.co/ibm-granite/granite-4.0-h-tiny/raw/main/README.md. Observed at 2026-10-07T20:24:02.984574+00:00.Wed, 07 Oct 2026 20:24:02 GMThttps://openmodeldrops.com/models/ibm-granite/granite-4.0-h-tiny/44931e87243650be105cLGAI-EXAONE/EXAONE-4.5-33B has author-reported evidence for "Reason through hard problems": Author reports WeMath 79.1 (EXAONE 4.5 33B (Reasoning)). Source: https://huggingface.co/LGAI-EXAONE/EXAONE-4.5-33B/raw/main/README.md. Observed at 2026-10-07T20:23:59.146267+00:00.Wed, 07 Oct 2026 20:23:59 GMThttps://openmodeldrops.com/models/LGAI-EXAONE/EXAONE-4.5-33B/edaeb83933e2f1da14bbgoogle/t5gemma-2-270m-270m has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 22.9 (few-shot). Source: https://huggingface.co/google/t5gemma-2-270m-270m/raw/main/README.md. Observed at 2026-10-07T20:23:53.868549+00:00.Wed, 07 Oct 2026 20:23:53 GMThttps://openmodeldrops.com/models/google/t5gemma-2-270m-270m/99a6397a04314578321fgoogle/gemma-3n-E2B-it-litert-lm has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 44.3 (E2B PT). Source: https://huggingface.co/google/gemma-3n-E2B-it-litert-lm/raw/main/README.md. Observed at 2026-10-07T20:23:51.160134+00:00.Wed, 07 Oct 2026 20:23:51 GMThttps://openmodeldrops.com/models/google/gemma-3n-E2B-it-litert-lm/bae29190f6e0807f2e19nvidia/Nemotron-Cascade-2-30B-A3B has author-reported evidence for "Reason through hard problems": Author reports IMO AnswerBench 79.3. Source: https://huggingface.co/nvidia/Nemotron-Cascade-2-30B-A3B/raw/main/README.md. Observed at 2026-10-07T20:23:50.705826+00:00.Wed, 07 Oct 2026 20:23:50 GMThttps://openmodeldrops.com/models/nvidia/Nemotron-Cascade-2-30B-A3B/0da57b51692fac1e7dd4NousResearch/Llama-2-13b-hf has author-reported evidence for "Reason through hard problems": Author reports Math 28.7 (Llama 2 13B). Source: https://huggingface.co/NousResearch/Llama-2-13b-hf/raw/main/README.md. Observed at 2026-10-07T20:23:47.454292+00:00.Wed, 07 Oct 2026 20:23:47 GMThttps://openmodeldrops.com/models/NousResearch/Llama-2-13b-hf/45040e0627a793d5d628moonshotai/Kimi-VL-A3B-Thinking-2506 has author-reported evidence for "Reason through hard problems": Author reports MMLU 82.0. Source: https://huggingface.co/moonshotai/Kimi-VL-A3B-Thinking-2506/raw/main/README.md. Observed at 2026-10-07T20:23:39.433875+00:00.Wed, 07 Oct 2026 20:23:39 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-VL-A3B-Thinking-2506/bd3f3d1985319d9b6868Nanbeige/Nanbeige4.2-3B has author-reported evidence for "Reason through hard problems": Author reports GPQA-Diamond 87.4. Source: https://huggingface.co/Nanbeige/Nanbeige4.2-3B/raw/main/README.md. Observed at 2026-10-07T20:23:38.903393+00:00.Wed, 07 Oct 2026 20:23:38 GMThttps://openmodeldrops.com/models/Nanbeige/Nanbeige4.2-3B/9ec3a56278024835d4e3deepseek-ai/DeepSeek-V2 has author-reported evidence for "Reason through hard problems": Author reports MMLU 78.5 (Base Model). Source: https://huggingface.co/deepseek-ai/DeepSeek-V2/raw/main/README.md. Observed at 2026-10-07T20:23:35.667092+00:00.Wed, 07 Oct 2026 20:23:35 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V2/347f5951bb9166b62bbdstepfun-ai/Step3-VL-10B has author-reported evidence for "Reason through hard problems": Author reports AIME 2025 94.43 (PaCoRe). Source: https://huggingface.co/stepfun-ai/Step3-VL-10B/raw/main/README.md. Observed at 2026-10-07T20:23:35.103901+00:00.Wed, 07 Oct 2026 20:23:35 GMThttps://openmodeldrops.com/models/stepfun-ai/Step3-VL-10B/843b85db8f6206f53dcbmoonshotai/Moonlight-16B-A3B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 70.0. Source: https://huggingface.co/moonshotai/Moonlight-16B-A3B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:23:30.562405+00:00.Wed, 07 Oct 2026 20:23:30 GMThttps://openmodeldrops.com/models/moonshotai/Moonlight-16B-A3B-Instruct/e8a990ed5f123a70c1c5allenai/OLMoE-1B-7B-0924 has author-reported evidence for "Reason through hard problems": Author reports MMLU 54.1 (OLMoE-1B-7B). Source: https://huggingface.co/allenai/OLMoE-1B-7B-0924/raw/main/README.md. Observed at 2026-10-07T20:23:27.868597+00:00.Wed, 07 Oct 2026 20:23:27 GMThttps://openmodeldrops.com/models/allenai/OLMoE-1B-7B-0924/3fb9264307d82e180fc3google/gemma-2-27b has author-reported evidence for "Reason through hard problems": Author reports MMLU 75.2 (5-shot, top-1). Source: https://huggingface.co/google/gemma-2-27b/raw/main/README.md. Observed at 2026-10-07T20:23:27.210852+00:00.Wed, 07 Oct 2026 20:23:27 GMThttps://openmodeldrops.com/models/google/gemma-2-27b/57534f7da3751f3e1fd2LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct has author-reported evidence for "Reason through hard problems": Author reports LiveBench 33.0 (EXAONE 3.5 2.4B). Source: https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:23:25.996657+00:00.Wed, 07 Oct 2026 20:23:25 GMThttps://openmodeldrops.com/models/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct/84202a13a026fe63e5aaai21labs/Jamba-v0.1 has author-reported evidence for "Reason through hard problems": Author reports MMLU 67.4%. Source: https://huggingface.co/ai21labs/Jamba-v0.1/raw/main/README.md. Observed at 2026-10-07T20:23:24.644284+00:00.Wed, 07 Oct 2026 20:23:24 GMThttps://openmodeldrops.com/models/ai21labs/Jamba-v0.1/1d792c7f91aa3602e492nvidia/Llama-3.1-Nemotron-Nano-8B-v1 has author-reported evidence for "Reason through hard problems": Author reports MATH500 95.4% (Reasoning On). Source: https://huggingface.co/nvidia/Llama-3.1-Nemotron-Nano-8B-v1/raw/main/README.md. Observed at 2026-10-07T20:23:23.648226+00:00.Wed, 07 Oct 2026 20:23:23 GMThttps://openmodeldrops.com/models/nvidia/Llama-3.1-Nemotron-Nano-8B-v1/4ec2bfed328afe5b7f5dNousResearch/Llama-2-7b-chat-hf has author-reported evidence for "Reason through hard problems": Author reports Math 14.6 (Llama 2 7B). Source: https://huggingface.co/NousResearch/Llama-2-7b-chat-hf/raw/main/README.md. Observed at 2026-10-07T20:23:23.481543+00:00.Wed, 07 Oct 2026 20:23:23 GMThttps://openmodeldrops.com/models/NousResearch/Llama-2-7b-chat-hf/325c2bdbb460e71426b2deepseek-ai/DeepSeek-V2-Chat has author-reported evidence for "Reason through hard problems": Author reports MMLU 78.5 (DeepSeek-V2 (MoE-236B)). Source: https://huggingface.co/deepseek-ai/DeepSeek-V2-Chat/raw/main/README.md. Observed at 2026-10-07T20:23:17.479579+00:00.Wed, 07 Oct 2026 20:23:17 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V2-Chat/eb0c0e928f92d3a042a3deepseek-ai/DeepSeek-V3.1-Base has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 91.8 (DeepSeek V3.1-NonThinking). Source: https://huggingface.co/deepseek-ai/DeepSeek-V3.1-Base/raw/main/README.md. Observed at 2026-10-07T20:23:14.346504+00:00.Wed, 07 Oct 2026 20:23:14 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V3.1-Base/fbb7a87c960c15793ebbornith-ai/Ornith-1.5-9B has author-reported evidence for "Reason through hard problems": Author reports GPQA Diamond 86.4. Source: https://huggingface.co/ornith-ai/Ornith-1.5-9B/raw/main/README.md. Observed at 2026-10-07T20:23:11.012299+00:00.Wed, 07 Oct 2026 20:23:11 GMThttps://openmodeldrops.com/models/ornith-ai/Ornith-1.5-9B/d6890a21321081d87e3dopenchat/openchat_3.5 has author-reported evidence for "Reason through hard problems": Author reports MMLU 64.3 (OpenChat 3.5). Source: https://huggingface.co/openchat/openchat_3.5/raw/main/README.md. Observed at 2026-10-07T20:23:07.660201+00:00.Wed, 07 Oct 2026 20:23:07 GMThttps://openmodeldrops.com/models/openchat/openchat_3.5/60409639cf7e936fdd31google/gemma-3n-E4B-it-litert-lm has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 52.9 (E4B PT). Source: https://huggingface.co/google/gemma-3n-E4B-it-litert-lm/raw/main/README.md. Observed at 2026-10-07T20:23:04.269735+00:00.Wed, 07 Oct 2026 20:23:04 GMThttps://openmodeldrops.com/models/google/gemma-3n-E4B-it-litert-lm/f5ccf21dd62daabec595ibm-granite/granite-4.1-3b has author-reported evidence for "Reason through hard problems": Author reports MMLU 67.02 (5-shot). Source: https://huggingface.co/ibm-granite/granite-4.1-3b/raw/main/README.md. Observed at 2026-10-07T20:23:00.041756+00:00.Wed, 07 Oct 2026 20:23:00 GMThttps://openmodeldrops.com/models/ibm-granite/granite-4.1-3b/882f98c360d7be30ce70NousResearch/Meta-Llama-3.1-8B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 69.4 (Instruction tuned models, 8B). Source: https://huggingface.co/NousResearch/Meta-Llama-3.1-8B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:22:53.929484+00:00.Wed, 07 Oct 2026 20:22:53 GMThttps://openmodeldrops.com/models/NousResearch/Meta-Llama-3.1-8B-Instruct/e85664f97b14f51d8de2microsoft/DialoGPT-small has author-reported evidence for "Reason through hard problems": Author reports MMLU (5-shot) 25.81. Source: https://huggingface.co/microsoft/DialoGPT-small/raw/main/README.md. Observed at 2026-10-07T20:22:46.495574+00:00.Wed, 07 Oct 2026 20:22:46 GMThttps://openmodeldrops.com/models/microsoft/DialoGPT-small/8c25db3e364fc403e20bgoogle/gemma-3-12b-pt has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 72.6 (Gemma 3 PT 12B). Source: https://huggingface.co/google/gemma-3-12b-pt/raw/main/README.md. Observed at 2026-10-07T20:22:41.760613+00:00.Wed, 07 Oct 2026 20:22:41 GMThttps://openmodeldrops.com/models/google/gemma-3-12b-pt/c4898b5f44cbfc512357google/gemma-4-26B-A4B-it-assistant has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 82.6% (Gemma 4 26B A4B). Source: https://huggingface.co/google/gemma-4-26B-A4B-it-assistant/raw/main/README.md. Observed at 2026-10-07T20:22:38.751487+00:00.Wed, 07 Oct 2026 20:22:38 GMThttps://openmodeldrops.com/models/google/gemma-4-26B-A4B-it-assistant/e1161d30ce70cde4d771nvidia/NVLM-D-72B has author-reported evidence for "Reason through hard problems": Author reports MMLU 82.0 (NVLM-D 1.0 72B (Megatron)). Source: https://huggingface.co/nvidia/NVLM-D-72B/raw/main/README.md. Observed at 2026-10-07T20:22:35.886125+00:00.Wed, 07 Oct 2026 20:22:35 GMThttps://openmodeldrops.com/models/nvidia/NVLM-D-72B/bc8b1766a6991ded55daHuggingFaceTB/SmolLM2-360M has author-reported evidence for "Reason through hard problems": Author reports MMLU (cloze) 35.8 (Base Pre-Trained Model). Source: https://huggingface.co/HuggingFaceTB/SmolLM2-360M/raw/main/README.md. Observed at 2026-10-07T20:22:30.881465+00:00.Wed, 07 Oct 2026 20:22:30 GMThttps://openmodeldrops.com/models/HuggingFaceTB/SmolLM2-360M/4cca6f3fa954671f82a2microsoft/Phi-4-mini-reasoning has author-reported evidence for "Reason through hard problems": Author reports AIME 57.5 (Phi-4-mini-reasoning (3.8B)). Source: https://huggingface.co/microsoft/Phi-4-mini-reasoning/raw/main/README.md. Observed at 2026-10-07T20:22:30.719617+00:00.Wed, 07 Oct 2026 20:22:30 GMThttps://openmodeldrops.com/models/microsoft/Phi-4-mini-reasoning/7f140040bdb28b7de983NousResearch/Llama-2-7b-hf has author-reported evidence for "Reason through hard problems": Author reports Math 14.6 (Llama 2 7B). Source: https://huggingface.co/NousResearch/Llama-2-7b-hf/raw/main/README.md. Observed at 2026-10-07T20:22:25.254586+00:00.Wed, 07 Oct 2026 20:22:25 GMThttps://openmodeldrops.com/models/NousResearch/Llama-2-7b-hf/5f9e78cd87d3f56f9aa5Qwen/Qwen2-1.5B has author-reported evidence for "Reason through hard problems": Author reports MMLU 56.5. Source: https://huggingface.co/Qwen/Qwen2-1.5B/raw/main/README.md. Observed at 2026-10-07T20:22:21.413447+00:00.Wed, 07 Oct 2026 20:22:21 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-1.5B/32ebf5ee168c5206b7eeQwen/Qwen-7B has author-reported evidence for "Reason through hard problems": Author reports MMLU 58.2 (5-shot). Source: https://huggingface.co/Qwen/Qwen-7B/raw/main/README.md. Observed at 2026-10-07T20:22:20.180963+00:00.Wed, 07 Oct 2026 20:22:20 GMThttps://openmodeldrops.com/models/Qwen/Qwen-7B/c3caffe787b7f6bbc91aLiquidAI/LFM2-1.2B has author-reported evidence for "Reason through hard problems": Author reports MMLU 55.23 (LFM2-1.2B). Source: https://huggingface.co/LiquidAI/LFM2-1.2B/raw/main/README.md. Observed at 2026-10-07T20:22:15.242882+00:00.Wed, 07 Oct 2026 20:22:15 GMThttps://openmodeldrops.com/models/LiquidAI/LFM2-1.2B/1745f66130770e35e168tencent/Hunyuan-A13B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 88.17 (Hunyuan-A13B). Source: https://huggingface.co/tencent/Hunyuan-A13B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:22:14.627728+00:00.Wed, 07 Oct 2026 20:22:14 GMThttps://openmodeldrops.com/models/tencent/Hunyuan-A13B-Instruct/7b089bf0943bfb82cc99allenai/Olmo-3-7B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MATH 87.3 (Olmo 3 Instruct 7B). Source: https://huggingface.co/allenai/Olmo-3-7B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:22:08.179979+00:00.Wed, 07 Oct 2026 20:22:08 GMThttps://openmodeldrops.com/models/allenai/Olmo-3-7B-Instruct/2bf8a530b9cd66c5f1cfQwen/Qwen2-7B has author-reported evidence for "Reason through hard problems": Author reports MMLU 70.3. Source: https://huggingface.co/Qwen/Qwen2-7B/raw/main/README.md. Observed at 2026-10-07T20:22:06.506492+00:00.Wed, 07 Oct 2026 20:22:06 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-7B/81502209be13ccbb4ca1microsoft/bitnet-b1.58-2B-4T has author-reported evidence for "Reason through hard problems": Author reports MMLU 53.17. Source: https://huggingface.co/microsoft/bitnet-b1.58-2B-4T/raw/main/README.md. Observed at 2026-10-07T20:22:03.760978+00:00.Wed, 07 Oct 2026 20:22:03 GMThttps://openmodeldrops.com/models/microsoft/bitnet-b1.58-2B-4T/6f6e3540cfdbbf510c72LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct has author-reported evidence for "Reason through hard problems": Author reports LiveBench 39.8 (EXAONE 3.5 7.8B). Source: https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:22:02.431577+00:00.Wed, 07 Oct 2026 20:22:02 GMThttps://openmodeldrops.com/models/LGAI-EXAONE/EXAONE-3.5-7.8B-Instruct/e605e9b8af9c5fdde547allenai/OLMo-2-0425-1B has author-reported evidence for "Reason through hard problems": Author reports BBH 35.0 (OLMo 2 1B). Source: https://huggingface.co/allenai/OLMo-2-0425-1B/raw/main/README.md. Observed at 2026-10-07T20:22:00.762458+00:00.Wed, 07 Oct 2026 20:22:00 GMThttps://openmodeldrops.com/models/allenai/OLMo-2-0425-1B/3486cfae690b9919300fEleutherAI/gpt-neo-1.3B has author-reported evidence for "Reason through hard problems": Author reports MathQA 24.05% (GPT-Neo 1.3B). Source: https://huggingface.co/EleutherAI/gpt-neo-1.3B/raw/main/README.md. Observed at 2026-10-07T20:21:56.000220+00:00.Wed, 07 Oct 2026 20:21:56 GMThttps://openmodeldrops.com/models/EleutherAI/gpt-neo-1.3B/4d7cef525efdc5e76557ibm-granite/granite-4.1-8b has author-reported evidence for "Reason through hard problems": Author reports MMLU 73.84 (8B Dense, 5-shot). Source: https://huggingface.co/ibm-granite/granite-4.1-8b/raw/main/README.md. Observed at 2026-10-07T20:21:54.760115+00:00.Wed, 07 Oct 2026 20:21:54 GMThttps://openmodeldrops.com/models/ibm-granite/granite-4.1-8b/511e3c928f0af393f418deepseek-ai/DeepSeek-V4-Flash-DSpark has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 89.4 (DeepSeek-V4-Flash-Base). Source: https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-DSpark/raw/main/README.md. Observed at 2026-10-07T20:21:51.774058+00:00.Wed, 07 Oct 2026 20:21:51 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V4-Flash-DSpark/8a892ea274aa567b1d12teknium/OpenHermes-2.5-Mistral-7B has author-reported evidence for "Reason through hard problems": Author reports agieval_sat_math 0.3773 (AGI-Eval Version 0). Source: https://huggingface.co/teknium/OpenHermes-2.5-Mistral-7B/raw/main/README.md. Observed at 2026-10-07T20:21:47.450559+00:00.Wed, 07 Oct 2026 20:21:47 GMThttps://openmodeldrops.com/models/teknium/OpenHermes-2.5-Mistral-7B/916ec88340d140749e4fgoogle/gemma-3-1b-pt has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 28.4 (Gemma 3 PT 1B, few-shot). Source: https://huggingface.co/google/gemma-3-1b-pt/raw/main/README.md. Observed at 2026-10-07T20:21:46.075480+00:00.Wed, 07 Oct 2026 20:21:46 GMThttps://openmodeldrops.com/models/google/gemma-3-1b-pt/c0432dc159efa57e3c32microsoft/Phi-3.5-MoE-instruct has author-reported evidence for "Reason through hard problems": Author reports BigBench Hard CoT (0-shot) 79.1. Source: https://huggingface.co/microsoft/Phi-3.5-MoE-instruct/raw/main/README.md. Observed at 2026-10-07T20:21:43.666035+00:00.Wed, 07 Oct 2026 20:21:43 GMThttps://openmodeldrops.com/models/microsoft/Phi-3.5-MoE-instruct/a667f91011fcbe5b956cgoogle/gemma-3-4b-pt has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 50.9 (Gemma 3 PT 4B). Source: https://huggingface.co/google/gemma-3-4b-pt/raw/main/README.md. Observed at 2026-10-07T20:21:40.164915+00:00.Wed, 07 Oct 2026 20:21:40 GMThttps://openmodeldrops.com/models/google/gemma-3-4b-pt/d22cd31cd9ea1d446bd0deepseek-ai/DeepSeek-V2-Lite-Chat has author-reported evidence for "Reason through hard problems": Author reports MMLU 58.3 (DeepSeek-V2-Lite (MoE-16B) Base). Source: https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite-Chat/raw/main/README.md. Observed at 2026-10-07T20:21:39.588871+00:00.Wed, 07 Oct 2026 20:21:39 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V2-Lite-Chat/aa2296792c8dd27fdfacdeepseek-ai/DeepSeek-V2-Lite has author-reported evidence for "Reason through hard problems": Author reports MMLU 58.3 (Base Model). Source: https://huggingface.co/deepseek-ai/DeepSeek-V2-Lite/raw/main/README.md. Observed at 2026-10-07T20:21:38.654457+00:00.Wed, 07 Oct 2026 20:21:38 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V2-Lite/11e2cc6d7ab03fd7763cEleutherAI/gpt-neo-2.7B has author-reported evidence for "Reason through hard problems": Author reports MathQA 24.72% (GPT-Neo 2.7B). Source: https://huggingface.co/EleutherAI/gpt-neo-2.7B/raw/main/README.md. Observed at 2026-10-07T20:21:35.749319+00:00.Wed, 07 Oct 2026 20:21:35 GMThttps://openmodeldrops.com/models/EleutherAI/gpt-neo-2.7B/7e2f9a67f3724a7769b3XiaomiMiMo/MiMo-V2-Flash has author-reported evidence for "Reason through hard problems": Author reports BBH 88.5 (Base, 3-shot). Source: https://huggingface.co/XiaomiMiMo/MiMo-V2-Flash/raw/main/README.md. Observed at 2026-10-07T20:21:29.045379+00:00.Wed, 07 Oct 2026 20:21:29 GMThttps://openmodeldrops.com/models/XiaomiMiMo/MiMo-V2-Flash/dd2252456143739e5b45apple/OpenELM-1_1B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 25.65 (LLM360). Source: https://huggingface.co/apple/OpenELM-1_1B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:21:28.157148+00:00.Wed, 07 Oct 2026 20:21:28 GMThttps://openmodeldrops.com/models/apple/OpenELM-1_1B-Instruct/43b6d1afcb956e393a1aQwen/Qwen-7B-Chat has author-reported evidence for "Reason through hard problems": Author reports MMLU 55.8 (BF16). Source: https://huggingface.co/Qwen/Qwen-7B-Chat/raw/main/README.md. Observed at 2026-10-07T20:21:24.891709+00:00.Wed, 07 Oct 2026 20:21:24 GMThttps://openmodeldrops.com/models/Qwen/Qwen-7B-Chat/0eabb963c4808b808ff8EleutherAI/gpt-neo-125m has author-reported evidence for "Reason through hard problems": Author reports MMLU (5-shot) 25.97. Source: https://huggingface.co/EleutherAI/gpt-neo-125m/raw/main/README.md. Observed at 2026-10-07T20:21:19.446803+00:00.Wed, 07 Oct 2026 20:21:19 GMThttps://openmodeldrops.com/models/EleutherAI/gpt-neo-125m/57e32609f59e2141f68cgoogle/gemma-4-31B-it-assistant has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 85.2% (Gemma 4 31B). Source: https://huggingface.co/google/gemma-4-31B-it-assistant/raw/main/README.md. Observed at 2026-10-07T20:21:17.344339+00:00.Wed, 07 Oct 2026 20:21:17 GMThttps://openmodeldrops.com/models/google/gemma-4-31B-it-assistant/a143a401f94426a9399eQwen/Qwen2-0.5B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 37.9 (Qwen2-0.5B-Instruct). Source: https://huggingface.co/Qwen/Qwen2-0.5B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:21:08.181247+00:00.Wed, 07 Oct 2026 20:21:08 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-0.5B-Instruct/f53c2225be1788634ef5Qwen/Qwen3-Next-80B-A3B-Thinking has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 82.7. Source: https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking/raw/main/README.md. Observed at 2026-10-07T20:21:08.022739+00:00.Wed, 07 Oct 2026 20:21:08 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-Next-80B-A3B-Thinking/6acf834cf2c5ae5b9cacgoogle/gemma-4-12B has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 77.2% (Gemma 4 12B Unified). Source: https://huggingface.co/google/gemma-4-12B/raw/main/README.md. Observed at 2026-10-07T20:21:05.712743+00:00.Wed, 07 Oct 2026 20:21:05 GMThttps://openmodeldrops.com/models/google/gemma-4-12B/3ccfcaa4d73b1cc5c463XiaomiMiMo/MiMo-V2.5-Pro has author-reported evidence for "Reason through hard problems": Author reports BBH 88.4 (3-shot). Source: https://huggingface.co/XiaomiMiMo/MiMo-V2.5-Pro/raw/main/README.md. Observed at 2026-10-07T20:21:04.477404+00:00.Wed, 07 Oct 2026 20:21:04 GMThttps://openmodeldrops.com/models/XiaomiMiMo/MiMo-V2.5-Pro/1544ca9927ede25360b8Qwen/Qwen2-0.5B has author-reported evidence for "Reason through hard problems": Author reports MMLU 45.4. Source: https://huggingface.co/Qwen/Qwen2-0.5B/raw/main/README.md. Observed at 2026-10-07T20:20:57.672631+00:00.Wed, 07 Oct 2026 20:20:57 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-0.5B/4dcd37804a91c941c65dQwen/Qwen3-30B-A3B-Thinking-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 80.9. Source: https://huggingface.co/Qwen/Qwen3-30B-A3B-Thinking-2507/raw/main/README.md. Observed at 2026-10-07T20:20:57.251386+00:00.Wed, 07 Oct 2026 20:20:57 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-30B-A3B-Thinking-2507/d9d5687bea0967b4dcb5EleutherAI/gpt-neox-20b has author-reported evidence for "Reason through hard problems": Author reports MMLU (5-shot) 25.0 (Open LLM Leaderboard). Source: https://huggingface.co/EleutherAI/gpt-neox-20b/raw/main/README.md. Observed at 2026-10-07T20:20:52.185481+00:00.Wed, 07 Oct 2026 20:20:52 GMThttps://openmodeldrops.com/models/EleutherAI/gpt-neox-20b/0020b3c843ed339300d3google/gemma-3-270m-it has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 26.7 (Gemma 3 IT 270m, few-shot). Source: https://huggingface.co/google/gemma-3-270m-it/raw/main/README.md. Observed at 2026-10-07T20:20:52.122791+00:00.Wed, 07 Oct 2026 20:20:52 GMThttps://openmodeldrops.com/models/google/gemma-3-270m-it/f50ffe8919cfedb09baemicrosoft/Phi-3-mini-128k-instruct has author-reported evidence for "Reason through hard problems": Author reports GPQA 29.7 (June 2024 Update). Source: https://huggingface.co/microsoft/Phi-3-mini-128k-instruct/raw/main/README.md. Observed at 2026-10-07T20:20:51.777947+00:00.Wed, 07 Oct 2026 20:20:51 GMThttps://openmodeldrops.com/models/microsoft/Phi-3-mini-128k-instruct/e20fa3b1fcce041c66bdgoogle/gemma-2-9b has author-reported evidence for "Reason through hard problems": Author reports MMLU 71.3 (5-shot, top-1). Source: https://huggingface.co/google/gemma-2-9b/raw/main/README.md. Observed at 2026-10-07T20:20:51.345666+00:00.Wed, 07 Oct 2026 20:20:51 GMThttps://openmodeldrops.com/models/google/gemma-2-9b/1b88d1d815c7c9b53c7eQwen/Qwen-72B has author-reported evidence for "Reason through hard problems": Author reports MMLU 77.4 (5-shot). Source: https://huggingface.co/Qwen/Qwen-72B/raw/main/README.md. Observed at 2026-10-07T20:20:47.134415+00:00.Wed, 07 Oct 2026 20:20:47 GMThttps://openmodeldrops.com/models/Qwen/Qwen-72B/35e550941528c90e1d80Qwen/Qwen2-1.5B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 52.4 (Qwen2-1.5B-Instruct). Source: https://huggingface.co/Qwen/Qwen2-1.5B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:20:42.883114+00:00.Wed, 07 Oct 2026 20:20:42 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-1.5B-Instruct/cb8a2b398ac75d0d45c2Qwen/Qwen2-72B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 82.3 (English). Source: https://huggingface.co/Qwen/Qwen2-72B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:20:32.699551+00:00.Wed, 07 Oct 2026 20:20:32 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-72B-Instruct/0d3293de770f13564febgoogle/gemma-4-E2B has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 60.0% (Gemma 4 E2B). Source: https://huggingface.co/google/gemma-4-E2B/raw/main/README.md. Observed at 2026-10-07T20:20:29.538589+00:00.Wed, 07 Oct 2026 20:20:29 GMThttps://openmodeldrops.com/models/google/gemma-4-E2B/f5495b3b52e2ffe66348deepseek-ai/DeepSeek-V3.2-Exp has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 85.0 (Reasoning Mode w/o Tool Use). Source: https://huggingface.co/deepseek-ai/DeepSeek-V3.2-Exp/raw/main/README.md. Observed at 2026-10-07T20:20:22.825669+00:00.Wed, 07 Oct 2026 20:20:22 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V3.2-Exp/0f97404bef97534b5ef0HuggingFaceTB/SmolLM3-3B has author-reported evidence for "Reason through hard problems": Author reports AIME 2025 36.7 (Instruction Model, Extended Thinking). Source: https://huggingface.co/HuggingFaceTB/SmolLM3-3B/raw/main/README.md. Observed at 2026-10-07T20:20:22.315403+00:00.Wed, 07 Oct 2026 20:20:22 GMThttps://openmodeldrops.com/models/HuggingFaceTB/SmolLM3-3B/252a0c34841c694d2bffmoonshotai/Kimi-Linear-48B-A3B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro (4k context length) 51.0. Source: https://huggingface.co/moonshotai/Kimi-Linear-48B-A3B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:20:20.525238+00:00.Wed, 07 Oct 2026 20:20:20 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-Linear-48B-A3B-Instruct/7f97fe7d7460f5d667c2google/gemma-4-26B-A4B has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 82.6% (Gemma 4 26B A4B). Source: https://huggingface.co/google/gemma-4-26B-A4B/raw/main/README.md. Observed at 2026-10-07T20:20:11.847069+00:00.Wed, 07 Oct 2026 20:20:11 GMThttps://openmodeldrops.com/models/google/gemma-4-26B-A4B/ac2aa4aee1688718b457HuggingFaceTB/SmolLM2-135M has author-reported evidence for "Reason through hard problems": Author reports MMLU (cloze) 31.5 (SmolLM2-135M-8k). Source: https://huggingface.co/HuggingFaceTB/SmolLM2-135M/raw/main/README.md. Observed at 2026-10-07T20:20:07.352337+00:00.Wed, 07 Oct 2026 20:20:07 GMThttps://openmodeldrops.com/models/HuggingFaceTB/SmolLM2-135M/638c58d2b1b4e722b90dNousResearch/Hermes-3-Llama-3.1-8B has author-reported evidence for "Reason through hard problems": Author reports MATH Lvl 5 (4-Shot) 4.76. Source: https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B/raw/main/README.md. Observed at 2026-10-07T20:20:05.430809+00:00.Wed, 07 Oct 2026 20:20:05 GMThttps://openmodeldrops.com/models/NousResearch/Hermes-3-Llama-3.1-8B/cafe4f1485b41e9d7aebQwen/Qwen3-235B-A22B-Thinking-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 84.4. Source: https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507/raw/main/README.md. Observed at 2026-10-07T20:20:04.754537+00:00.Wed, 07 Oct 2026 20:20:04 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-235B-A22B-Thinking-2507/7e8d9e5743e75f86f7f3microsoft/Phi-4-mini-instruct has author-reported evidence for "Reason through hard problems": Author reports BigBench Hard (0-shot, CoT) 70.4 (Phi-4 mini-Ins). Source: https://huggingface.co/microsoft/Phi-4-mini-instruct/raw/main/README.md. Observed at 2026-10-07T20:19:56.578522+00:00.Wed, 07 Oct 2026 20:19:56 GMThttps://openmodeldrops.com/models/microsoft/Phi-4-mini-instruct/03e3e621e8b471abae03google/gemma-4-E4B has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 69.4% (Gemma 4 E4B). Source: https://huggingface.co/google/gemma-4-E4B/raw/main/README.md. Observed at 2026-10-07T20:19:55.830091+00:00.Wed, 07 Oct 2026 20:19:55 GMThttps://openmodeldrops.com/models/google/gemma-4-E4B/e9550f00c5576af61005nvidia/NVIDIA-Nemotron-Nano-9B-v2 has author-reported evidence for "Reason through hard problems": Author reports AIME25 72.1% (Reasoning On). Source: https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-9B-v2/raw/main/README.md. Observed at 2026-10-07T20:19:53.646336+00:00.Wed, 07 Oct 2026 20:19:53 GMThttps://openmodeldrops.com/models/nvidia/NVIDIA-Nemotron-Nano-9B-v2/e93711afff147d3e63bfgoogle/gemma-3-270m has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 26.7 (Gemma 3 IT 270m (few-shot)). Source: https://huggingface.co/google/gemma-3-270m/raw/main/README.md. Observed at 2026-10-07T20:19:53.643758+00:00.Wed, 07 Oct 2026 20:19:53 GMThttps://openmodeldrops.com/models/google/gemma-3-270m/902dd02d7302c7cf9965microsoft/Phi-3-mini-4k-instruct has author-reported evidence for "Reason through hard problems": Author reports GPQA 30.6 (June 2024 Update). Source: https://huggingface.co/microsoft/Phi-3-mini-4k-instruct/raw/main/README.md. Observed at 2026-10-07T20:19:50.357769+00:00.Wed, 07 Oct 2026 20:19:50 GMThttps://openmodeldrops.com/models/microsoft/Phi-3-mini-4k-instruct/61ae71d2eba63f07ae03google/diffusiongemma-26B-A4B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 77.6% (DiffusionGemma 26B A4B). Source: https://huggingface.co/google/diffusiongemma-26B-A4B-it/raw/main/README.md. Observed at 2026-10-07T20:19:48.067859+00:00.Wed, 07 Oct 2026 20:19:48 GMThttps://openmodeldrops.com/models/google/diffusiongemma-26B-A4B-it/75368068810ed20ba097Qwen/Qwen3-4B-Thinking-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 74.0 (Qwen3-4B-Thinking-2507). Source: https://huggingface.co/Qwen/Qwen3-4B-Thinking-2507/raw/main/README.md. Observed at 2026-10-07T20:19:47.128419+00:00.Wed, 07 Oct 2026 20:19:47 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-4B-Thinking-2507/dcf1a0f3f62a97693b58moonshotai/Kimi-K2-Thinking has author-reported evidence for "Reason through hard problems": Author reports HLE (Text-only) 51.0 (heavy). Source: https://huggingface.co/moonshotai/Kimi-K2-Thinking/raw/main/README.md. Observed at 2026-10-07T20:19:33.970202+00:00.Wed, 07 Oct 2026 20:19:33 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-K2-Thinking/90c2d0dd715d47a52db8deepseek-ai/DeepSeek-R1-Distill-Llama-70B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 pass@1 70.0 (DeepSeek-R1-Distill-Llama-70B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B/raw/main/README.md. Observed at 2026-10-07T20:19:28.506663+00:00.Wed, 07 Oct 2026 20:19:28 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Llama-70B/ac81cf55c21646cc1b26stepfun-ai/Step-3.5-Flash has author-reported evidence for "Reason through hard problems": Author reports AIME 2025 97.3. Source: https://huggingface.co/stepfun-ai/Step-3.5-Flash/raw/main/README.md. Observed at 2026-10-07T20:19:26.201489+00:00.Wed, 07 Oct 2026 20:19:26 GMThttps://openmodeldrops.com/models/stepfun-ai/Step-3.5-Flash/2eceb64b471d206b125egoogle/gemma-4-31B has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 85.2% (Gemma 4 31B). Source: https://huggingface.co/google/gemma-4-31B/raw/main/README.md. Observed at 2026-10-07T20:19:22.456254+00:00.Wed, 07 Oct 2026 20:19:22 GMThttps://openmodeldrops.com/models/google/gemma-4-31B/28125f4fe48ba76eb766google/gemma-2-2b has author-reported evidence for "Reason through hard problems": Author reports MMLU 51.3 (5-shot, top-1). Source: https://huggingface.co/google/gemma-2-2b/raw/main/README.md. Observed at 2026-10-07T20:19:18.173147+00:00.Wed, 07 Oct 2026 20:19:18 GMThttps://openmodeldrops.com/models/google/gemma-2-2b/380a637583979d6e1122deepseek-ai/DeepSeek-R1-0528-Qwen3-8B has author-reported evidence for "Reason through hard problems": Author reports AIME 24 86.0 (DeepSeek-R1-0528-Qwen3-8B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B/raw/main/README.md. Observed at 2026-10-07T20:19:13.226996+00:00.Wed, 07 Oct 2026 20:19:13 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B/75cc4e4e137c06ab877cmeta-models/Muse-Glimmer-30B has author-reported evidence for "Reason through hard problems": Author reports AIME 2026 94.7 (High Reasoning). Source: https://huggingface.co/meta-models/Muse-Glimmer-30B/raw/main/README.md. Observed at 2026-10-07T20:19:12.040914+00:00.Wed, 07 Oct 2026 20:19:12 GMThttps://openmodeldrops.com/models/meta-models/Muse-Glimmer-30B/de2ea8b547fe3bb36e96thinkingmachines/Inkling has author-reported evidence for "Reason through hard problems": Author reports HLE (text only) 29.7%. Source: https://huggingface.co/thinkingmachines/Inkling/raw/main/README.md. Observed at 2026-10-07T20:19:05.702369+00:00.Wed, 07 Oct 2026 20:19:05 GMThttps://openmodeldrops.com/models/thinkingmachines/Inkling/6575ac192bde7d55ffe2Qwen/Qwen3.5-0.8B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 42.3 (Knowledge & STEM (Thinking)). Source: https://huggingface.co/Qwen/Qwen3.5-0.8B/raw/main/README.md. Observed at 2026-10-07T20:18:48.986805+00:00.Wed, 07 Oct 2026 20:18:48 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-0.8B/050c4ff6040100f706b1Qwen/Qwen2-7B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 70.5 (English). Source: https://huggingface.co/Qwen/Qwen2-7B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:18:32.724291+00:00.Wed, 07 Oct 2026 20:18:32 GMThttps://openmodeldrops.com/models/Qwen/Qwen2-7B-Instruct/7ead40f3f9ef46dd1decQwen/Qwen3.5-2B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 66.5 (Knowledge & STEM (Thinking)). Source: https://huggingface.co/Qwen/Qwen3.5-2B/raw/main/README.md. Observed at 2026-10-07T20:18:30.029124+00:00.Wed, 07 Oct 2026 20:18:30 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-2B/e6e55d550cc4886108fcQwen/Qwen2.5-VL-32B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU 78.4 (Qwen2.5-VL-32B). Source: https://huggingface.co/Qwen/Qwen2.5-VL-32B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:18:28.397309+00:00.Wed, 07 Oct 2026 20:18:28 GMThttps://openmodeldrops.com/models/Qwen/Qwen2.5-VL-32B-Instruct/92e2359bd1a570443c54deepseek-ai/DeepSeek-R1-Distill-Llama-8B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 cons@64 80.0 (DeepSeek-R1-Distill-Llama-8B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-8B/raw/main/README.md. Observed at 2026-10-07T20:18:26.119294+00:00.Wed, 07 Oct 2026 20:18:26 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Llama-8B/ef027ae9f99a1ed55e2bgoogle/gemma-4-E2B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 60.0% (Gemma 4 E2B). Source: https://huggingface.co/google/gemma-4-E2B-it/raw/main/README.md. Observed at 2026-10-07T20:18:23.721758+00:00.Wed, 07 Oct 2026 20:18:23 GMThttps://openmodeldrops.com/models/google/gemma-4-E2B-it/6250c4c18dabcc84b2c7deepseek-ai/DeepSeek-R1-Distill-Qwen-14B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 cons@64 80.0 (DeepSeek-R1-Distill-Qwen-14B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B/raw/main/README.md. Observed at 2026-10-07T20:18:21.256145+00:00.Wed, 07 Oct 2026 20:18:21 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B/dba954739decae3d3b94google/gemma-4-12B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 77.2% (Gemma 4 12B Unified). Source: https://huggingface.co/google/gemma-4-12B-it/raw/main/README.md. Observed at 2026-10-07T20:18:18.754019+00:00.Wed, 07 Oct 2026 20:18:18 GMThttps://openmodeldrops.com/models/google/gemma-4-12B-it/70e29fb8120c7378b314Qwen/Qwen3-30B-A3B-Instruct-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 78.4. Source: https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507/raw/main/README.md. Observed at 2026-10-07T20:18:16.108331+00:00.Wed, 07 Oct 2026 20:18:16 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-30B-A3B-Instruct-2507/930175b0b17af9a34ae1deepseek-ai/DeepSeek-R1-Distill-Qwen-7B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 cons@64 83.3 (DeepSeek-R1-Distill-Qwen-7B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B/raw/main/README.md. Observed at 2026-10-07T20:18:14.246508+00:00.Wed, 07 Oct 2026 20:18:14 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B/fcf68b5a1b95e96b8070Qwen/Qwen3.5-122B-A10B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 86.7 (Qwen3.5-122B-A10B). Source: https://huggingface.co/Qwen/Qwen3.5-122B-A10B/raw/main/README.md. Observed at 2026-10-07T20:18:08.754187+00:00.Wed, 07 Oct 2026 20:18:08 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-122B-A10B/f664e28e84fb875a2ebddeepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 cons@64 52.7 (DeepSeek-R1-Distill-Qwen-1.5B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B/raw/main/README.md. Observed at 2026-10-07T20:17:54.056144+00:00.Wed, 07 Oct 2026 20:17:54 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B/3884f9e8ade2b3577359zai-org/GLM-4.7 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 84.3. Source: https://huggingface.co/zai-org/GLM-4.7/raw/main/README.md. Observed at 2026-10-07T20:17:53.256404+00:00.Wed, 07 Oct 2026 20:17:53 GMThttps://openmodeldrops.com/models/zai-org/GLM-4.7/58e3546fd957fe02b16agoogle/gemma-3-1b-it has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 28.4 (Gemma 3 PT 1B). Source: https://huggingface.co/google/gemma-3-1b-it/raw/main/README.md. Observed at 2026-10-07T20:17:48.055197+00:00.Wed, 07 Oct 2026 20:17:48 GMThttps://openmodeldrops.com/models/google/gemma-3-1b-it/5d63497cc4232db3435cdeepseek-ai/DeepSeek-R1-Distill-Qwen-32B has author-reported evidence for "Reason through hard problems": Author reports AIME 2024 cons@64 83.3 (DeepSeek-R1-Distill-Qwen-32B). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B/raw/main/README.md. Observed at 2026-10-07T20:17:45.604036+00:00.Wed, 07 Oct 2026 20:17:45 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B/0040dbe6b3fc4da03353Qwen/Qwen3.5-4B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 79.1 (Qwen3.5-4B). Source: https://huggingface.co/Qwen/Qwen3.5-4B/raw/main/README.md. Observed at 2026-10-07T20:17:40.399321+00:00.Wed, 07 Oct 2026 20:17:40 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-4B/5ec9998651366692a627microsoft/phi-4 has author-reported evidence for "Reason through hard problems": Author reports MMLU 84.8 (phi-4 (14B)). Source: https://huggingface.co/microsoft/phi-4/raw/main/README.md. Observed at 2026-10-07T20:17:30.565050+00:00.Wed, 07 Oct 2026 20:17:30 GMThttps://openmodeldrops.com/models/microsoft/phi-4/bbdfda39d3b137d4dbadzai-org/GLM-5.1 has author-reported evidence for "Reason through hard problems": Author reports HLE 31.0. Source: https://huggingface.co/zai-org/GLM-5.1/raw/main/README.md. Observed at 2026-10-07T20:17:29.948373+00:00.Wed, 07 Oct 2026 20:17:29 GMThttps://openmodeldrops.com/models/zai-org/GLM-5.1/1392b2daa2ac1ac71cebgoogle/gemma-3-12b-it has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 72.6 (Gemma 3 PT 12B). Source: https://huggingface.co/google/gemma-3-12b-it/raw/main/README.md. Observed at 2026-10-07T20:17:29.551896+00:00.Wed, 07 Oct 2026 20:17:29 GMThttps://openmodeldrops.com/models/google/gemma-3-12b-it/a603b51e619cbe8439fcgoogle/gemma-4-E4B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 69.4% (Gemma 4 E4B). Source: https://huggingface.co/google/gemma-4-E4B-it/raw/main/README.md. Observed at 2026-10-07T20:17:16.087842+00:00.Wed, 07 Oct 2026 20:17:16 GMThttps://openmodeldrops.com/models/google/gemma-4-E4B-it/6544487a1ae2c802eb12Qwen/Qwen3-235B-A22B-Instruct-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 83.0. Source: https://huggingface.co/Qwen/Qwen3-235B-A22B-Instruct-2507/raw/main/README.md. Observed at 2026-10-07T20:17:13.932469+00:00.Wed, 07 Oct 2026 20:17:13 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-235B-A22B-Instruct-2507/b0d7ee53b236c0a9a004Qwen/Qwen3-4B-Instruct-2507 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 69.6. Source: https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507/raw/main/README.md. Observed at 2026-10-07T20:17:09.502463+00:00.Wed, 07 Oct 2026 20:17:09 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-4B-Instruct-2507/1f1499f85101507da2d1deepseek-ai/DeepSeek-V3 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 89.1 (Chat Model). Source: https://huggingface.co/deepseek-ai/DeepSeek-V3/raw/main/README.md. Observed at 2026-10-07T20:17:04.835528+00:00.Wed, 07 Oct 2026 20:17:04 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V3/ca3aab73e7d5b01bf0b9Qwen/Qwen3.5-27B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 86.1 (Qwen3.5-27B). Source: https://huggingface.co/Qwen/Qwen3.5-27B/raw/main/README.md. Observed at 2026-10-07T20:17:03.923445+00:00.Wed, 07 Oct 2026 20:17:03 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-27B/eaf1292dd68f82b6e249zai-org/GLM-5 has author-reported evidence for "Reason through hard problems": Author reports HLE 30.5. Source: https://huggingface.co/zai-org/GLM-5/raw/main/README.md. Observed at 2026-10-07T20:17:02.190452+00:00.Wed, 07 Oct 2026 20:17:02 GMThttps://openmodeldrops.com/models/zai-org/GLM-5/ef85c1bc85853428aa76Qwen/Qwen3.5-35B-A3B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 85.3 (Qwen3.5-35B-A3B). Source: https://huggingface.co/Qwen/Qwen3.5-35B-A3B/raw/main/README.md. Observed at 2026-10-07T20:16:54.817315+00:00.Wed, 07 Oct 2026 20:16:54 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-35B-A3B/32f0ae352eeae7caf06cdeepseek-ai/DeepSeek-R1-0528 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 93.4 (DeepSeek R1 0528). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1-0528/raw/main/README.md. Observed at 2026-10-07T20:16:50.849233+00:00.Wed, 07 Oct 2026 20:16:50 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1-0528/1389f311075549ac4b0fmoonshotai/Kimi-K2.5 has author-reported evidence for "Reason through hard problems": Author reports AIME 2025 96.1 (Thinking). Source: https://huggingface.co/moonshotai/Kimi-K2.5/raw/main/README.md. Observed at 2026-10-07T20:16:46.965549+00:00.Wed, 07 Oct 2026 20:16:46 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-K2.5/9cab26cba25ed57caf30google/gemma-3-27b-it has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 77.7 (Gemma 3 PT 27B). Source: https://huggingface.co/google/gemma-3-27b-it/raw/main/README.md. Observed at 2026-10-07T20:16:45.849742+00:00.Wed, 07 Oct 2026 20:16:45 GMThttps://openmodeldrops.com/models/google/gemma-3-27b-it/396798a31f24f9e583d7Qwen/Qwen3-Next-80B-A3B-Instruct has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 80.6. Source: https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Instruct/raw/main/README.md. Observed at 2026-10-07T20:16:39.354303+00:00.Wed, 07 Oct 2026 20:16:39 GMThttps://openmodeldrops.com/models/Qwen/Qwen3-Next-80B-A3B-Instruct/371e2836eb9df1a14a9bgoogle/gemma-3-4b-it has author-reported evidence for "Reason through hard problems": Author reports BIG-Bench Hard 50.9 (Gemma 3 PT 4B). Source: https://huggingface.co/google/gemma-3-4b-it/raw/main/README.md. Observed at 2026-10-07T20:16:22.438688+00:00.Wed, 07 Oct 2026 20:16:22 GMThttps://openmodeldrops.com/models/google/gemma-3-4b-it/82ca6c3d847fea4cc4dbQwen/Qwen3.5-397B-A17B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 87.8. Source: https://huggingface.co/Qwen/Qwen3.5-397B-A17B/raw/main/README.md. Observed at 2026-10-07T20:16:22.298169+00:00.Wed, 07 Oct 2026 20:16:22 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-397B-A17B/aee79a42f3430ca17a14deepseek-ai/DeepSeek-V4-Flash has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 89.4 (DeepSeek-V4-Flash-Base). Source: https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash/raw/main/README.md. Observed at 2026-10-07T20:16:15.080747+00:00.Wed, 07 Oct 2026 20:16:15 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V4-Flash/5f80594d268b779da903zai-org/GLM-4.7-Flash has author-reported evidence for "Reason through hard problems": Author reports AIME 25 91.6. Source: https://huggingface.co/zai-org/GLM-4.7-Flash/raw/main/README.md. Observed at 2026-10-07T20:16:08.438286+00:00.Wed, 07 Oct 2026 20:16:08 GMThttps://openmodeldrops.com/models/zai-org/GLM-4.7-Flash/e2df21ebffe13f43696fdeepseek-ai/DeepSeek-R1 has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 92.9 (DeepSeek R1). Source: https://huggingface.co/deepseek-ai/DeepSeek-R1/raw/main/README.md. Observed at 2026-10-07T20:15:57.018851+00:00.Wed, 07 Oct 2026 20:15:57 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-R1/02ea6780cf28224e9d0cdeepseek-ai/DeepSeek-V4-Pro has author-reported evidence for "Reason through hard problems": Author reports MMLU-Redux (EM) 90.8 (DeepSeek-V4-Pro-Base). Source: https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/raw/main/README.md. Observed at 2026-10-07T20:15:50.350090+00:00.Wed, 07 Oct 2026 20:15:50 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V4-Pro/bef3a06da17072fce629deepseek-ai/DeepSeek-V4.1-Flash has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro (EM) 74.1 (DeepSeek-V4.1-Flash-Base). Source: https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash/raw/main/README.md. Observed at 2026-10-07T20:15:43.203420+00:00.Wed, 07 Oct 2026 20:15:43 GMThttps://openmodeldrops.com/models/deepseek-ai/DeepSeek-V4.1-Flash/feb1a2fbdafb5c67dae1Qwen/Qwen3.5-9B has author-reported evidence for "Reason through hard problems": Author reports MMLU-Pro 82.5 (Qwen3.5-9B). Source: https://huggingface.co/Qwen/Qwen3.5-9B/raw/main/README.md. Observed at 2026-10-07T20:15:39.011526+00:00.Wed, 07 Oct 2026 20:15:39 GMThttps://openmodeldrops.com/models/Qwen/Qwen3.5-9B/773b040c5dd57c3a9a5emoonshotai/Kimi-K3 has author-reported evidence for "Reason through hard problems": Author reports GPQA Diamond 93.5. Source: https://huggingface.co/moonshotai/Kimi-K3/raw/main/README.md. Observed at 2026-10-07T20:15:39.011373+00:00.Wed, 07 Oct 2026 20:15:39 GMThttps://openmodeldrops.com/models/moonshotai/Kimi-K3/7759fbeba0c7d3de7bf0google/gemma-4-31B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 85.2% (Gemma 4 31B). Source: https://huggingface.co/google/gemma-4-31B-it/raw/main/README.md. Observed at 2026-10-07T20:15:39.011299+00:00.Wed, 07 Oct 2026 20:15:39 GMThttps://openmodeldrops.com/models/google/gemma-4-31B-it/91651d22f71c2324cf59google/gemma-4-26B-A4B-it has author-reported evidence for "Reason through hard problems": Author reports MMLU Pro 82.6% (Gemma 4 26B A4B). Source: https://huggingface.co/google/gemma-4-26B-A4B-it/raw/main/README.md. Observed at 2026-10-07T20:14:26.358568+00:00.Wed, 07 Oct 2026 20:14:26 GMThttps://openmodeldrops.com/models/google/gemma-4-26B-A4B-it/88b045378a5040f09c4bzai-org/GLM-5.2 has author-reported evidence for "Reason through hard problems": Author reports HLE 40.5. Source: https://huggingface.co/zai-org/GLM-5.2/raw/main/README.md. Observed at 2026-10-07T20:14:26.358517+00:00.Wed, 07 Oct 2026 20:14:26 GMThttps://openmodeldrops.com/models/zai-org/GLM-5.2/bd4cbfa581420f0053c7