models.dev · 多厂商固定目录
成功
上次成功：2026-09-30T02:15:05.886Z
DATA & PROVENANCE
每一项神通，都有依据
目录、服务报价与公开评测分别核验。所有页面读取同一组已完成的发布快照。
评测覆盖 161 / 426 个型号；缺测不计零分。目录、报价与来源成绩分别统计；评测覆盖含独立列出的历史题集和配置，不代表均有当前全科成绩。同步状态与全部缺口 →
最近一次更新
2026-09-30T02:15:05.886Z · 全部已启用来源成功
LiteLLM · 提供方报价与容量
成功
上次成功：2026-09-30T02:15:05.886Z
Vercel AI Gateway · 服务目录
成功
上次成功：2026-09-30T02:15:05.886Z
OpenRouter · 路由参考目录
成功
上次成功：2026-09-30T02:15:05.886Z
Hugging Face · 官方模型仓库元数据
成功
上次成功：2026-09-30T02:15:05.886Z
livebench
成功
上次成功：2026-09-30T02:15:05.886Z
Epoch AI · 官方指数与自有成绩
成功
上次成功：2026-09-30T02:15:05.886Z
Arena · 官方 Hugging Face 数据集
成功
上次成功：2026-09-30T02:15:05.886Z
计划任务入口已具备；当前没有启用无人值守调度。上游失败不会把旧时间改成新的成功时间。
426已收录型号
670服务端点
本批范围：models.dev 登记材料中的 45 家厂商型号
采集截点 2026/9/30 10:15:05（北京时间）
发布时可比评测覆盖：0 / 426（独立 BenchmarkRun）；来源发布榜覆盖另列，流行度未接入。
文本参考成本最低 0 USD：输入 10,000 + 输出 1,000 tokens；范围为 423 个完整报价端点。最低报价来自 North Mini Code / cohere、Gemma 4 26B A4B IT / google · litellm、Gemma 4 31B IT / google · litellm、Gemma 3 27B IT / google · litellm。
models.dev · 多厂商固定目录
Copyright (c) 2025 models.dev — MIT License；本站仅摘取已核验的目录规格及标准文本报价。
当前发布采用的采集时间：2026-09-30T02:15:05.886Z
状态：新鲜期内。新鲜期 7 天、有效期 30 天；材料复核到期须重新审核。
MIT ↗具体材料（632 项）
- models/openai/gpt-4.1.toml ↗
- openai/models/gpt-4.1.toml ↗
- models/openai/gpt-4.1-mini.toml ↗
- openai/models/gpt-4.1-mini.toml ↗
- models/openai/gpt-4.1-nano.toml ↗
- openai/models/gpt-4.1-nano.toml ↗
- models/ai21/jamba-large.toml ↗
- ai21/models/jamba-large.toml ↗
- models/ai21/jamba-mini.toml ↗
- ai21/models/jamba-mini.toml ↗
- models/aisingapore/gemma-sea-lion-v4-27b-it.toml ↗
- models/alibaba/qwen-flash.toml ↗
- alibaba/models/qwen-flash.toml ↗
- models/alibaba/qwen-image-2.1.toml ↗
- models/alibaba/qwen-max.toml ↗
- alibaba/models/qwen-max.toml ↗
- models/alibaba/qwen-omni-turbo.toml ↗
- alibaba/models/qwen-omni-turbo.toml ↗
- models/alibaba/qwen-plus.toml ↗
- alibaba/models/qwen-plus.toml ↗
- models/alibaba/qwen-turbo.toml ↗
- alibaba/models/qwen-turbo.toml ↗
- models/alibaba/qwen-vl-max.toml ↗
- alibaba/models/qwen-vl-max.toml ↗
- models/alibaba/qwen-vl-plus.toml ↗
- alibaba/models/qwen-vl-plus.toml ↗
- models/alibaba/qwen2-5-vl-72b-instruct.toml ↗
- alibaba/models/qwen2-5-vl-72b-instruct.toml ↗
- models/alibaba/qwen2.5-coder-0.5b.toml ↗
- models/alibaba/qwen2.5-coder-32b-instruct.toml ↗
- models/alibaba/qwen3-235b-a22b-instruct-2507.toml ↗
- models/alibaba/qwen3-235b-a22b.toml ↗
- alibaba/models/qwen3-235b-a22b.toml ↗
- models/alibaba/qwen3-30b-a3b.toml ↗
- models/alibaba/qwen3-32b.toml ↗
- alibaba/models/qwen3-32b.toml ↗
- models/alibaba/qwen3-coder-30b-a3b-instruct.toml ↗
- alibaba/models/qwen3-coder-30b-a3b-instruct.toml ↗
- models/alibaba/qwen3-coder-480b-a35b-instruct.toml ↗
- alibaba/models/qwen3-coder-480b-a35b-instruct.toml ↗
- models/alibaba/qwen3-coder-flash.toml ↗
- alibaba/models/qwen3-coder-flash.toml ↗
- models/alibaba/qwen3-coder-next.toml ↗
- models/alibaba/qwen3-coder-plus.toml ↗
- alibaba/models/qwen3-coder-plus.toml ↗
- models/alibaba/qwen3-max.toml ↗
- alibaba/models/qwen3-max.toml ↗
- models/alibaba/qwen3-next-80b-a3b-instruct.toml ↗
- alibaba/models/qwen3-next-80b-a3b-instruct.toml ↗
- models/alibaba/qwen3-next-80b-a3b-thinking.toml ↗
- alibaba/models/qwen3-next-80b-a3b-thinking.toml ↗
- models/alibaba/qwen3-vl-235b-a22b-instruct.toml ↗
- models/alibaba/qwen3-vl-235b-a22b-thinking.toml ↗
- models/alibaba/qwen3-vl-plus.toml ↗
- alibaba/models/qwen3-vl-plus.toml ↗
- models/alibaba/qwen3.5-122b-a10b.toml ↗
- alibaba/models/qwen3.5-122b-a10b.toml ↗
- models/alibaba/qwen3.5-27b.toml ↗
- alibaba/models/qwen3.5-27b.toml ↗
- models/alibaba/qwen3.5-35b-a3b.toml ↗
- alibaba/models/qwen3.5-35b-a3b.toml ↗
- models/alibaba/qwen3.5-397b-a17b.toml ↗
- alibaba/models/qwen3.5-397b-a17b.toml ↗
- models/alibaba/qwen3.5-9b.toml ↗
- models/alibaba/qwen3.5-flash.toml ↗
- alibaba/models/qwen3.5-flash.toml ↗
- models/alibaba/qwen3.5-plus.toml ↗
- alibaba/models/qwen3.5-plus.toml ↗
- models/alibaba/qwen3.6-27b.toml ↗
- alibaba/models/qwen3.6-27b.toml ↗
- models/alibaba/qwen3.6-35b-a3b.toml ↗
- alibaba/models/qwen3.6-35b-a3b.toml ↗
- models/alibaba/qwen3.6-flash.toml ↗
- alibaba/models/qwen3.6-flash.toml ↗
- models/alibaba/qwen3.6-max-preview.toml ↗
- alibaba/models/qwen3.6-max-preview.toml ↗
- models/alibaba/qwen3.6-plus.toml ↗
- alibaba/models/qwen3.6-plus.toml ↗
- models/alibaba/qwen3.7-flash.toml ↗
- alibaba/models/qwen3.7-flash.toml ↗
- models/alibaba/qwen3.7-max.toml ↗
- alibaba/models/qwen3.7-max.toml ↗
- models/alibaba/qwen3.7-plus.toml ↗
- alibaba/models/qwen3.7-plus.toml ↗
- models/alibaba/qwen3.8-2.4t-a95b.toml ↗
- models/alibaba/qwen3.8-27b.toml ↗
- models/alibaba/qwen3.8-flash-next.toml ↗
- models/alibaba/qwen3.8-flash.toml ↗
- alibaba/models/qwen3.8-flash.toml ↗
- models/alibaba/qwen3.8-max-0902.toml ↗
- models/alibaba/qwen3.8-max-preview.toml ↗
- models/alibaba/qwen3.8-max-prime.toml ↗
- models/alibaba/qwen3.8-max.toml ↗
- alibaba/models/qwen3.8-max.toml ↗
- models/alibaba/qwen3.8-omni-flash.toml ↗
- alibaba/models/qwen3.8-omni-flash.toml ↗
- models/alibaba/qwq-32b.toml ↗
- models/alibaba/qwq-plus.toml ↗
- alibaba/models/qwq-plus.toml ↗
- models/amazon/nova-2-lite.toml ↗
- models/amazon/nova-lite.toml ↗
- models/amazon/nova-micro.toml ↗
- models/amazon/nova-premier.toml ↗
- models/amazon/nova-pro.toml ↗
- models/anthropic/claude-3-5-haiku-20241022.toml ↗
- models/anthropic/claude-3-5-sonnet-20241022.toml ↗
- models/anthropic/claude-3-7-sonnet-20250219.toml ↗
- models/anthropic/claude-3-haiku-20240307.toml ↗
- models/anthropic/claude-fable-5-1.toml ↗
- anthropic/models/claude-fable-5-1.toml ↗
- models/anthropic/claude-fable-5.toml ↗
- anthropic/models/claude-fable-5.toml ↗
- models/anthropic/claude-haiku-4-5-20251001.toml ↗
- anthropic/models/claude-haiku-4-5-20251001.toml ↗
- models/anthropic/claude-haiku-4-5.toml ↗
- anthropic/models/claude-haiku-4-5.toml ↗
- models/anthropic/claude-mythos-5.toml ↗
- models/anthropic/claude-opus-4-0.toml ↗
- models/anthropic/claude-opus-4-1-20250805.toml ↗
- models/anthropic/claude-opus-4-1.toml ↗
- models/anthropic/claude-opus-4-20250514.toml ↗
- models/anthropic/claude-opus-4-5-20251101.toml ↗
- anthropic/models/claude-opus-4-5-20251101.toml ↗
- models/anthropic/claude-opus-4-5.toml ↗
- anthropic/models/claude-opus-4-5.toml ↗
- models/anthropic/claude-opus-4-6.toml ↗
- anthropic/models/claude-opus-4-6.toml ↗
- models/anthropic/claude-opus-4-7.toml ↗
- anthropic/models/claude-opus-4-7.toml ↗
- models/anthropic/claude-opus-4-8.toml ↗
- anthropic/models/claude-opus-4-8.toml ↗
- models/anthropic/claude-opus-5-5.toml ↗
- anthropic/models/claude-opus-5-5.toml ↗
- models/anthropic/claude-opus-5.toml ↗
- anthropic/models/claude-opus-5.toml ↗
- models/anthropic/claude-sonnet-4-0.toml ↗
- models/anthropic/claude-sonnet-4-20250514.toml ↗
- models/anthropic/claude-sonnet-4-5-20250929.toml ↗
- anthropic/models/claude-sonnet-4-5-20250929.toml ↗
- models/anthropic/claude-sonnet-4-5.toml ↗
- anthropic/models/claude-sonnet-4-5.toml ↗
- models/anthropic/claude-sonnet-4-6.toml ↗
- anthropic/models/claude-sonnet-4-6.toml ↗
- models/anthropic/claude-sonnet-5-5.toml ↗
- anthropic/models/claude-sonnet-5-5.toml ↗
- models/anthropic/claude-sonnet-5.toml ↗
- anthropic/models/claude-sonnet-5.toml ↗
- models/arcee-ai/trinity-large-preview.toml ↗
- models/arcee-ai/trinity-large-thinking.toml ↗
- models/arcee-ai/trinity-mini.toml ↗
- models/arcee-ai/trinity-nano-preview.toml ↗
- models/bytedance-seed/seed-1-6-flash.toml ↗
- models/bytedance-seed/seed-1-6-vision.toml ↗
- models/bytedance-seed/seed-1-6.toml ↗
- models/bytedance-seed/seed-1-8.toml ↗
- models/bytedance-seed/seed-2.0-code.toml ↗
- models/bytedance-seed/seed-2.0-lite.toml ↗
- models/bytedance-seed/seed-2.0-mini.toml ↗
- models/bytedance-seed/seed-2.0-pro.toml ↗
- models/bytedance-seed/seed-2.1-pro.toml ↗
- models/bytedance-seed/seed-2.1-turbo.toml ↗
- models/bytedance-seed/seed-character.toml ↗
- models/bytedance-seed/seed-evolving.toml ↗
- models/cohere/c4ai-aya-expanse-32b.toml ↗
- cohere/models/c4ai-aya-expanse-32b.toml ↗
- models/cohere/c4ai-aya-expanse-8b.toml ↗
- cohere/models/c4ai-aya-expanse-8b.toml ↗
- models/cohere/c4ai-aya-vision-32b.toml ↗
- cohere/models/c4ai-aya-vision-32b.toml ↗
- models/cohere/c4ai-aya-vision-8b.toml ↗
- cohere/models/c4ai-aya-vision-8b.toml ↗
- models/cohere/command-a-03-2025.toml ↗
- cohere/models/command-a-03-2025.toml ↗
- models/cohere/command-a-plus-05-2026.toml ↗
- cohere/models/command-a-plus-05-2026.toml ↗
- models/cohere/command-a-reasoning-08-2025.toml ↗
- cohere/models/command-a-reasoning-08-2025.toml ↗
- models/cohere/command-a-translate-08-2025.toml ↗
- cohere/models/command-a-translate-08-2025.toml ↗
- models/cohere/command-a-vision-07-2025.toml ↗
- cohere/models/command-a-vision-07-2025.toml ↗
- models/cohere/command-r-08-2024.toml ↗
- cohere/models/command-r-08-2024.toml ↗
- models/cohere/command-r-plus-08-2024.toml ↗
- cohere/models/command-r-plus-08-2024.toml ↗
- models/cohere/command-r7b-12-2024.toml ↗
- cohere/models/command-r7b-12-2024.toml ↗
- models/cohere/command-r7b-arabic-02-2025.toml ↗
- cohere/models/command-r7b-arabic-02-2025.toml ↗
- models/cohere/north-mini-code-1-0.toml ↗
- cohere/models/north-mini-code-1-0.toml ↗
- models/deepreinforce/ornith-1.0-31b.toml ↗
- models/deepreinforce/ornith-1.0-35b.toml ↗
- models/deepreinforce/ornith-1.0-397b.toml ↗
- models/deepreinforce/ornith-1.0-9b.toml ↗
- models/deepreinforce/ornith-1.5-35b-a3b.toml ↗
- models/deepseek/deepseek-chat.toml ↗
- models/deepseek/deepseek-ocr-2.toml ↗
- models/deepseek/deepseek-r1-distill-qwen-32b.toml ↗
- models/deepseek/deepseek-r1.toml ↗
- models/deepseek/deepseek-reasoner.toml ↗
- models/deepseek/deepseek-v3-0324.toml ↗
- models/deepseek/deepseek-v3.1.toml ↗
- models/deepseek/deepseek-v3.2.toml ↗
- models/deepseek/deepseek-v3.toml ↗
- models/deepseek/deepseek-v4-flash-0423.toml ↗
- models/deepseek/deepseek-v4-flash-0731.toml ↗
- models/deepseek/deepseek-v4-flash-vision-exp.toml ↗
- models/deepseek/deepseek-v4-flash.toml ↗
- models/deepseek/deepseek-v4-pro-0423.toml ↗
- models/deepseek/deepseek-v4-pro-0813.toml ↗
- models/deepseek/deepseek-v4-pro.toml ↗
- models/deepseek/deepseek-v4.1-flash.toml ↗
- models/google/deep-research-max-preview-04-2026.toml ↗
- google/models/deep-research-max-preview-04-2026.toml ↗
- models/google/deep-research-preview-04-2026.toml ↗
- google/models/deep-research-preview-04-2026.toml ↗
- models/google/gemini-2.0-flash-lite.toml ↗
- models/google/gemini-2.0-flash.toml ↗
- models/google/gemini-2.5-computer-use-preview-10-2025.toml ↗
- google/models/gemini-2.5-computer-use-preview-10-2025.toml ↗
- models/google/gemini-2.5-flash-image.toml ↗
- google/models/gemini-2.5-flash-image.toml ↗
- models/google/gemini-2.5-flash-lite.toml ↗
- google/models/gemini-2.5-flash-lite.toml ↗
- models/google/gemini-2.5-flash-tts.toml ↗
- models/google/gemini-2.5-flash.toml ↗
- google/models/gemini-2.5-flash.toml ↗
- models/google/gemini-2.5-pro-tts.toml ↗
- models/google/gemini-2.5-pro.toml ↗
- google/models/gemini-2.5-pro.toml ↗
- models/google/gemini-3-flash-preview.toml ↗
- google/models/gemini-3-flash-preview.toml ↗
- models/google/gemini-3-pro-image-preview.toml ↗
- google/models/gemini-3-pro-image-preview.toml ↗
- models/google/gemini-3-pro-image.toml ↗
- google/models/gemini-3-pro-image.toml ↗
- models/google/gemini-3-pro-preview.toml ↗
- models/google/gemini-3.1-flash-image-preview.toml ↗
- models/google/gemini-3.1-flash-image.toml ↗
- google/models/gemini-3.1-flash-image.toml ↗
- models/google/gemini-3.1-flash-lite-image.toml ↗
- google/models/gemini-3.1-flash-lite-image.toml ↗
- models/google/gemini-3.1-flash-lite-preview.toml ↗
- google/models/gemini-3.1-flash-lite-preview.toml ↗
- models/google/gemini-3.1-flash-lite.toml ↗
- google/models/gemini-3.1-flash-lite.toml ↗
- models/google/gemini-3.1-flash-live-preview.toml ↗
- google/models/gemini-3.1-flash-live-preview.toml ↗
- models/google/gemini-3.1-flash-tts-preview.toml ↗
- google/models/gemini-3.1-flash-tts-preview.toml ↗
- models/google/gemini-3.1-pro-preview-customtools.toml ↗
- google/models/gemini-3.1-pro-preview-customtools.toml ↗
- models/google/gemini-3.1-pro-preview.toml ↗
- google/models/gemini-3.1-pro-preview.toml ↗
- models/google/gemini-3.5-flash-lite.toml ↗
- google/models/gemini-3.5-flash-lite.toml ↗
- models/google/gemini-3.5-flash.toml ↗
- google/models/gemini-3.5-flash.toml ↗
- models/google/gemini-3.5-live-translate-preview.toml ↗
- google/models/gemini-3.5-live-translate-preview.toml ↗
- models/google/gemini-3.5-transcribe-live.toml ↗
- models/google/gemini-3.6-flash.toml ↗
- google/models/gemini-3.6-flash.toml ↗
- models/google/gemini-3.7-flash.toml ↗
- google/models/gemini-3.7-flash.toml ↗
- models/google/gemini-3.8-flash.toml ↗
- google/models/gemini-3.8-flash.toml ↗
- models/google/gemini-embedding-001.toml ↗
- google/models/gemini-embedding-001.toml ↗
- models/google/gemini-embedding-2.toml ↗
- google/models/gemini-embedding-2.toml ↗
- models/google/gemini-flash-latest.toml ↗
- google/models/gemini-flash-latest.toml ↗
- models/google/gemini-flash-lite-latest.toml ↗
- google/models/gemini-flash-lite-latest.toml ↗
- models/google/gemini-omni-flash-preview.toml ↗
- google/models/gemini-omni-flash-preview.toml ↗
- models/google/gemini-robotics-er-1.6-preview.toml ↗
- models/google/gemma-3-12b-it.toml ↗
- models/google/gemma-3-27b-it.toml ↗
- models/google/gemma-3-4b-it.toml ↗
- models/google/gemma-4-12b-it.toml ↗
- models/google/gemma-4-26b-a4b-it.toml ↗
- google/models/gemma-4-26b-a4b-it.toml ↗
- models/google/gemma-4-31b-it.toml ↗
- google/models/gemma-4-31b-it.toml ↗
- models/google/lyria-3-clip-preview.toml ↗
- google/models/lyria-3-clip-preview.toml ↗
- models/google/lyria-3-pro-preview.toml ↗
- google/models/lyria-3-pro-preview.toml ↗
- models/google/veo-3.1-fast-generate-preview.toml ↗
- google/models/veo-3.1-fast-generate-preview.toml ↗
- models/google/veo-3.1-generate-preview.toml ↗
- google/models/veo-3.1-generate-preview.toml ↗
- models/google/veo-3.1-lite-generate-preview.toml ↗
- google/models/veo-3.1-lite-generate-preview.toml ↗
- models/ibm/granite-4-h-micro.toml ↗
- models/ibm/granite-4-h-small.toml ↗
- models/inclusionai/ling-3.0-flash-fin.toml ↗
- models/meituan/longcat-2.0.toml ↗
- models/meituan/longcat-2.5-preview.toml ↗
- models/meta/llama-3.1-70b-instruct.toml ↗
- models/meta/llama-3.1-8b-instruct.toml ↗
- models/meta/llama-3.2-11b-vision-instruct.toml ↗
- models/meta/llama-3.2-1b.toml ↗
- models/meta/llama-3.2-3b.toml ↗
- models/meta/llama-3.3-70b-instruct.toml ↗
- models/meta/llama-4-maverick-17b-instruct.toml ↗
- models/meta/llama-4-scout-17b-instruct.toml ↗
- models/meta/llama-guard-3-8b.toml ↗
- models/meta/muse-glimmer-30b.toml ↗
- models/meta/muse-spark-1.1.toml ↗
- meta/models/muse-spark-1.1.toml ↗
- models/meta/muse-spark-1.2.toml ↗
- meta/models/muse-spark-1.2.toml ↗
- models/meta/muse-spark-1.3.toml ↗
- meta/models/muse-spark-1.3.toml ↗
- models/microsoft/mai-code-1-flash.toml ↗
- models/microsoft/mai-code-1.1-flash.toml ↗
- models/microsoft/phi-4-mini.toml ↗
- models/minimax/image-01.toml ↗
- models/mistral/codestral-22b-v0.1.toml ↗
- models/mistral/codestral-latest.toml ↗
- mistral/models/codestral-latest.toml ↗
- models/mistral/devstral-2512.toml ↗
- mistral/models/devstral-2512.toml ↗
- models/mistral/devstral-medium-2507.toml ↗
- mistral/models/devstral-medium-2507.toml ↗
- models/mistral/devstral-medium-latest.toml ↗
- mistral/models/devstral-medium-latest.toml ↗
- models/mistral/devstral-small-2.toml ↗
- models/mistral/devstral-small-2507.toml ↗
- mistral/models/devstral-small-2507.toml ↗
- models/mistral/magistral-medium-latest.toml ↗
- mistral/models/magistral-medium-latest.toml ↗
- models/mistral/magistral-small-2506.toml ↗
- models/mistral/magistral-small-2509.toml ↗
- models/mistral/ministral-14b.toml ↗
- models/mistral/ministral-3-14b-instruct-2512.toml ↗
- models/mistral/ministral-3-3b-instruct-2512.toml ↗
- models/mistral/ministral-3-8b-instruct-2512.toml ↗
- models/mistral/ministral-3b.toml ↗
- models/mistral/ministral-8b-instruct-2410.toml ↗
- models/mistral/mistral-embed.toml ↗
- mistral/models/mistral-embed.toml ↗
- models/mistral/mistral-large-2411.toml ↗
- mistral/models/mistral-large-2411.toml ↗
- models/mistral/mistral-large-2512.toml ↗
- mistral/models/mistral-large-2512.toml ↗
- models/mistral/mistral-large-latest.toml ↗
- mistral/models/mistral-large-latest.toml ↗
- models/mistral/mistral-medium-2505.toml ↗
- mistral/models/mistral-medium-2505.toml ↗
- models/mistral/mistral-medium-2604.toml ↗
- mistral/models/mistral-medium-2604.toml ↗
- models/mistral/mistral-medium-latest.toml ↗
- mistral/models/mistral-medium-latest.toml ↗
- models/mistral/mistral-nemo.toml ↗
- mistral/models/mistral-nemo.toml ↗
- models/mistral/mistral-small-2506.toml ↗
- mistral/models/mistral-small-2506.toml ↗
- models/mistral/mistral-small-2603.toml ↗
- mistral/models/mistral-small-2603.toml ↗
- models/mistral/mistral-small-3-1-24b-instruct-2503.toml ↗
- models/mistral/mistral-small-latest.toml ↗
- mistral/models/mistral-small-latest.toml ↗
- models/mistral/pixtral-12b.toml ↗
- mistral/models/pixtral-12b.toml ↗
- models/mistral/pixtral-large-2502.toml ↗
- models/mistral/pixtral-large-latest.toml ↗
- mistral/models/pixtral-large-latest.toml ↗
- models/mistral/voxtral-mini-3b-2507.toml ↗
- models/mistral/voxtral-small-24b-2507.toml ↗
- models/mistral/voxtral-small-latest.toml ↗
- mistral/models/voxtral-small-latest.toml ↗
- models/mixedbread/toast-1.toml ↗
- models/moonshotai/kimi-k2-thinking-turbo.toml ↗
- models/moonshotai/kimi-k2-thinking.toml ↗
- models/moonshotai/kimi-k2.5.toml ↗
- models/moonshotai/kimi-k2.6.toml ↗
- moonshotai/models/kimi-k2.6.toml ↗
- models/moonshotai/kimi-k2.7-code-highspeed.toml ↗
- models/moonshotai/kimi-k2.7-code.toml ↗
- moonshotai/models/kimi-k2.7-code.toml ↗
- models/moonshotai/kimi-k2.8-preview.toml ↗
- models/moonshotai/kimi-k3.toml ↗
- moonshotai/models/kimi-k3.toml ↗
- models/motif-technologies/motif-3.toml ↗
- models/nex-agi/nex-n2-pro.toml ↗
- models/nvidia/llama-3.1-nemotron-70b-instruct.toml ↗
- models/nvidia/llama-3.1-nemotron-safety-guard-8b-v3.toml ↗
- models/nvidia/llama-3.1-nemotron-ultra-253b.toml ↗
- models/nvidia/llama-3.3-nemotron-super-49b-v1.5.toml ↗
- models/nvidia/llama-3.3-nemotron-super-49b-v1.toml ↗
- models/nvidia/llama-nemotron-embed-vl-1b-v2.toml ↗
- models/nvidia/llama-nemotron-rerank-vl-1b-v2.toml ↗
- models/nvidia/mistral-nemotron.toml ↗
- models/nvidia/nemotron-3-content-safety.toml ↗
- models/nvidia/nemotron-3-nano-30b-a3b.toml ↗
- models/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning.toml ↗
- models/nvidia/nemotron-3-super-120b-a12b.toml ↗
- models/nvidia/nemotron-3-ultra-550b-a55b.toml ↗
- models/nvidia/nemotron-3.5-content-safety.toml ↗
- models/nvidia/nemotron-3.5-lightning.toml ↗
- models/nvidia/nemotron-cascade-2-30b-a3b.toml ↗
- models/nvidia/nemotron-content-safety-reasoning-4b.toml ↗
- models/nvidia/nemotron-mini-4b-instruct.toml ↗
- models/nvidia/nemotron-nano-12b-v2-vl.toml ↗
- models/nvidia/nemotron-nano-9b-v2.toml ↗
- models/nvidia/nemotron-voicechat.toml ↗
- models/openai/gpt-3.5-turbo.toml ↗
- openai/models/gpt-3.5-turbo.toml ↗
- models/openai/gpt-4-turbo.toml ↗
- openai/models/gpt-4-turbo.toml ↗
- models/openai/gpt-4.toml ↗
- openai/models/gpt-4.toml ↗
- models/openai/gpt-4o-2024-05-13.toml ↗
- openai/models/gpt-4o-2024-05-13.toml ↗
- models/openai/gpt-4o-2024-08-06.toml ↗
- openai/models/gpt-4o-2024-08-06.toml ↗
- models/openai/gpt-4o-2024-11-20.toml ↗
- openai/models/gpt-4o-2024-11-20.toml ↗
- models/openai/gpt-4o-mini.toml ↗
- openai/models/gpt-4o-mini.toml ↗
- models/openai/gpt-4o.toml ↗
- openai/models/gpt-4o.toml ↗
- models/openai/gpt-5-chat-latest.toml ↗
- models/openai/gpt-5-codex.toml ↗
- models/openai/gpt-5-mini.toml ↗
- openai/models/gpt-5-mini.toml ↗
- models/openai/gpt-5-nano.toml ↗
- openai/models/gpt-5-nano.toml ↗
- models/openai/gpt-5-pro.toml ↗
- openai/models/gpt-5-pro.toml ↗
- models/openai/gpt-5.1-chat-latest.toml ↗
- models/openai/gpt-5.1-codex-max.toml ↗
- models/openai/gpt-5.1-codex-mini.toml ↗
- models/openai/gpt-5.1-codex.toml ↗
- models/openai/gpt-5.1.toml ↗
- openai/models/gpt-5.1.toml ↗
- models/openai/gpt-5.2-chat-latest.toml ↗
- openai/models/gpt-5.2-chat-latest.toml ↗
- models/openai/gpt-5.2-codex.toml ↗
- models/openai/gpt-5.2-pro.toml ↗
- openai/models/gpt-5.2-pro.toml ↗
- models/openai/gpt-5.2.toml ↗
- openai/models/gpt-5.2.toml ↗
- models/openai/gpt-5.3-chat-latest.toml ↗
- openai/models/gpt-5.3-chat-latest.toml ↗
- models/openai/gpt-5.3-codex-spark.toml ↗
- openai/models/gpt-5.3-codex-spark.toml ↗
- models/openai/gpt-5.3-codex.toml ↗
- openai/models/gpt-5.3-codex.toml ↗
- models/openai/gpt-5.4-mini.toml ↗
- openai/models/gpt-5.4-mini.toml ↗
- models/openai/gpt-5.4-nano.toml ↗
- openai/models/gpt-5.4-nano.toml ↗
- models/openai/gpt-5.4-pro.toml ↗
- openai/models/gpt-5.4-pro.toml ↗
- models/openai/gpt-5.4.toml ↗
- openai/models/gpt-5.4.toml ↗
- models/openai/gpt-5.5-instant.toml ↗
- models/openai/gpt-5.5-pro.toml ↗
- openai/models/gpt-5.5-pro.toml ↗
- models/openai/gpt-5.5.toml ↗
- openai/models/gpt-5.5.toml ↗
- models/openai/gpt-5.6-cyber.toml ↗
- models/openai/gpt-5.6-luna.toml ↗
- openai/models/gpt-5.6-luna.toml ↗
- models/openai/gpt-5.6-sol.toml ↗
- openai/models/gpt-5.6-sol.toml ↗
- models/openai/gpt-5.6-terra.toml ↗
- openai/models/gpt-5.6-terra.toml ↗
- models/openai/gpt-5.toml ↗
- openai/models/gpt-5.toml ↗
- models/openai/gpt-6-astra-fast.toml ↗
- models/openai/gpt-6-astra.toml ↗
- openai/models/gpt-6-astra.toml ↗
- models/openai/gpt-6-luna.toml ↗
- openai/models/gpt-6-luna.toml ↗
- models/openai/gpt-6-sol.toml ↗
- openai/models/gpt-6-sol.toml ↗
- models/openai/gpt-image-1.5.toml ↗
- openai/models/gpt-image-1.5.toml ↗
- models/openai/gpt-image-1.toml ↗
- openai/models/gpt-image-1.toml ↗
- models/openai/gpt-image-2.5-flare.toml ↗
- models/openai/gpt-image-2.5-sunburst.toml ↗
- models/openai/gpt-image-2.toml ↗
- openai/models/gpt-image-2.toml ↗
- models/openai/gpt-oss-120b.toml ↗
- models/openai/gpt-oss-20b.toml ↗
- models/openai/gpt-oss-safeguard-120b.toml ↗
- models/openai/gpt-oss-safeguard-20b.toml ↗
- models/openai/gpt-realtime-2.1.toml ↗
- openai/models/gpt-realtime-2.1.toml ↗
- models/openai/gpt-realtime-whisper.toml ↗
- models/openai/o1-pro.toml ↗
- openai/models/o1-pro.toml ↗
- models/openai/o1.toml ↗
- openai/models/o1.toml ↗
- models/openai/o3-deep-research.toml ↗
- models/openai/o3-mini.toml ↗
- openai/models/o3-mini.toml ↗
- models/openai/o3-pro.toml ↗
- openai/models/o3-pro.toml ↗
- models/openai/o3.toml ↗
- openai/models/o3.toml ↗
- models/openai/o4-mini-deep-research.toml ↗
- models/openai/o4-mini.toml ↗
- openai/models/o4-mini.toml ↗
- models/openai/whisper-large-v3-turbo.toml ↗
- models/openai/whisper-large-v3.toml ↗
- models/openbmb/minicpm5-1b.toml ↗
- models/openbmb/minicpm5-2b.toml ↗
- models/perplexity/sonar-deep-research.toml ↗
- models/perplexity/sonar-pro.toml ↗
- perplexity/models/sonar-pro.toml ↗
- models/perplexity/sonar-reasoning-pro.toml ↗
- perplexity/models/sonar-reasoning-pro.toml ↗
- models/perplexity/sonar.toml ↗
- perplexity/models/sonar.toml ↗
- models/poolside/laguna-m.1.toml ↗
- models/poolside/laguna-s-2.1.toml ↗
- models/poolside/laguna-xs-2.1.toml ↗
- models/poolside/laguna-xs.2.toml ↗
- models/quiverai/arrow-2-telos.toml ↗
- models/quiverai/arrow-2.toml ↗
- models/sakana/fugu-ultra.toml ↗
- sakana/models/fugu-ultra.toml ↗
- models/sakana/fugu.toml ↗
- sakana/models/fugu.toml ↗
- models/sakana/sakana-namazu.toml ↗
- sakana/models/sakana-namazu.toml ↗
- models/sarvam/sarvam-105b.toml ↗
- models/sarvam/sarvam-30b.toml ↗
- models/sdaia/allam-2-7b.toml ↗
- models/stepfun/step-3.5-flash-2603.toml ↗
- stepfun/models/step-3.5-flash-2603.toml ↗
- models/stepfun/step-3.5-flash.toml ↗
- stepfun/models/step-3.5-flash.toml ↗
- models/stepfun/step-3.7-flash.toml ↗
- stepfun/models/step-3.7-flash.toml ↗
- models/stepfun/step-5-preview.toml ↗
- stepfun/models/step-5-preview.toml ↗
- models/swiss-ai/apertus-70b.toml ↗
- models/swiss-ai/apertus-8b.toml ↗
- models/tencent/hy3-preview.toml ↗
- models/tencent/hy3.toml ↗
- models/tencent/hy4-preview.toml ↗
- models/thinkingmachines/inkling-small.toml ↗
- models/thinkingmachines/inkling.toml ↗
- models/trendyol/asure-12b.toml ↗
- models/typesafe/jev-latest.toml ↗
- models/unbiased/pareto.toml ↗
- models/upstage/solar-pro2.toml ↗
- models/upstage/solar-pro3.toml ↗
- models/upstage/solar-pro4.toml ↗
- upstage/models/solar-pro4.toml ↗
- models/vispark/vision-large.toml ↗
- models/vispark/vision-medium.toml ↗
- models/vispark/vision-small.toml ↗
- models/vivgrid/viv-fast.toml ↗
- vivgrid/models/viv-fast.toml ↗
- models/writer/palmyra-x4.toml ↗
- models/writer/palmyra-x5.toml ↗
- models/xai/grok-4.1-fast-reasoning.toml ↗
- models/xai/grok-4.1-fast.toml ↗
- models/xai/grok-4.20-0309-non-reasoning.toml ↗
- xai/models/grok-4.20-0309-non-reasoning.toml ↗
- models/xai/grok-4.20-0309-reasoning.toml ↗
- xai/models/grok-4.20-0309-reasoning.toml ↗
- models/xai/grok-4.3.toml ↗
- xai/models/grok-4.3.toml ↗
- models/xai/grok-4.5.toml ↗
- xai/models/grok-4.5.toml ↗
- models/xai/grok-4.6.toml ↗
- xai/models/grok-4.6.toml ↗
- models/xai/grok-4.7.toml ↗
- xai/models/grok-4.7.toml ↗
- models/xai/grok-build-0.1.toml ↗
- xai/models/grok-build-0.1.toml ↗
- models/xai/grok-imagine-image-2.0.toml ↗
- models/xai/grok-imagine-image-quality.toml ↗
- xai/models/grok-imagine-image-quality.toml ↗
- models/xai/grok-imagine-video-1.5.toml ↗
- xai/models/grok-imagine-video-1.5.toml ↗
- models/xai/grok-voice-stt-1.0.toml ↗
- models/xai/grok-voice-tts-1.0.toml ↗
- models/xiaomi/mimo-v2-flash.toml ↗
- xiaomi/models/mimo-v2-flash.toml ↗
- models/xiaomi/mimo-v2-omni.toml ↗
- xiaomi/models/mimo-v2-omni.toml ↗
- models/xiaomi/mimo-v2-pro.toml ↗
- xiaomi/models/mimo-v2-pro.toml ↗
- models/xiaomi/mimo-v2.5-pro-ultraspeed.toml ↗
- xiaomi/models/mimo-v2.5-pro-ultraspeed.toml ↗
- models/xiaomi/mimo-v2.5-pro.toml ↗
- xiaomi/models/mimo-v2.5-pro.toml ↗
- models/xiaomi/mimo-v2.5.toml ↗
- xiaomi/models/mimo-v2.5.toml ↗
- models/xiaomi/mimo-v2.6-flash.toml ↗
- xiaomi/models/mimo-v2.6-flash.toml ↗
- models/xiaomi/mimo-v2.6-pro-ultraspeed.toml ↗
- xiaomi/models/mimo-v2.6-pro-ultraspeed.toml ↗
- models/xiaomi/mimo-v2.6-pro.toml ↗
- xiaomi/models/mimo-v2.6-pro.toml ↗
- models/zhipuai/glm-5-turbo.toml ↗
- models/zhipuai/glm-5.1.toml ↗
- zhipuai/models/glm-5.1.toml ↗
- models/zhipuai/glm-5.2.toml ↗
- zhipuai/models/glm-5.2.toml ↗
- models/zhipuai/glm-5.3-flash.toml ↗
- zhipuai/models/glm-5.3-flash.toml ↗
- models/zhipuai/glm-5.3.toml ↗
- zhipuai/models/glm-5.3.toml ↗
- models/zhipuai/glm-5.toml ↗
- zhipuai/models/glm-5.toml ↗
- models/zhipuai/glm-5v-turbo.toml ↗
- zhipuai/models/glm-5v-turbo.toml ↗
- models/zhipuai/glm-image.toml ↗
- models/zhipuai/glm-4.5-air.toml ↗
- models/zhipuai/glm-4.5-flash.toml ↗
- models/zhipuai/glm-4.5.toml ↗
- models/zhipuai/glm-4.5v.toml ↗
- models/zhipuai/glm-4.6.toml ↗
- models/zhipuai/glm-4.6v-flash.toml ↗
- models/zhipuai/glm-4.6v.toml ↗
- models/zhipuai/glm-4.7-flash.toml ↗
- models/zhipuai/glm-4.7-flashx.toml ↗
- models/zhipuai/glm-4.7.toml ↗
LiteLLM · 提供方报价与容量
LiteLLM · 提供方报价与容量；本站仅提取型号、服务容量与数值报价；归属映射和单位转换由本站完成。
当前发布采用的采集时间：2026-09-30T02:15:05.886Z
状态：新鲜期内。新鲜期 7 天、有效期 30 天；材料复核到期须重新审核。
MIT ↗Vercel AI Gateway · 服务目录
Vercel AI Gateway · 服务目录；本站仅提取型号、服务容量与数值报价；归属映射和单位转换由本站完成。
当前发布采用的采集时间：2026-09-30T02:15:05.886Z
状态：新鲜期内。新鲜期 7 天、有效期 30 天；材料复核到期须重新审核。
公开 API 事实摘录；不声称开放数据许可 ↗具体材料（1 项）
OpenRouter · 路由参考目录
OpenRouter · 路由参考目录；本站仅提取型号、服务容量与数值报价；归属映射和单位转换由本站完成。
当前发布采用的采集时间：2026-09-30T02:15:05.886Z
状态：新鲜期内。新鲜期 7 天、有效期 30 天；材料复核到期须重新审核。
公开 API 事实摘录；不声称开放数据许可 ↗具体材料（1 项）
Hugging Face · 官方模型仓库元数据
Hugging Face 与各模型仓库发布者；只摘录许可证标识和 safetensors 总参数量，不复制模型卡或权重。
当前发布采用的采集时间：2026-09-30T02:15:05.886Z
状态：新鲜期内。新鲜期 7 天、有效期 30 天；材料复核到期须重新审核。
模型仓库公开元数据事实；模型许可证逐仓库保存 ↗具体材料（24 项）
- models/moonshotai/Kimi-K2.7-Code ↗
- models/deepseek-ai/DeepSeek-V3 ↗
- models/openai/gpt-oss-120b ↗
- models/nvidia/Llama-3.1-Nemotron-70B-Instruct ↗
- models/Qwen/QwQ-32B ↗
- models/google/gemma-3-27b-it ↗
- models/google/gemma-3-12b-it ↗
- models/google/gemma-3-4b-it ↗
- models/Qwen/Qwen3-32B ↗
- models/google/gemma-4-31B-it ↗
- models/Qwen/Qwen3-Coder-480B-A35B-Instruct ↗
- models/Qwen/Qwen3-Next-80B-A3B-Instruct ↗
- models/Qwen/Qwen3-Next-80B-A3B-Thinking ↗
- models/moonshotai/Kimi-K3 ↗
- models/deepseek-ai/DeepSeek-V4-Pro ↗
- models/mistralai/Devstral-2-123B-Instruct-2512 ↗
- models/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-NVFP4 ↗
- models/Qwen/Qwen3.6-27B ↗
- models/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-NVFP4 ↗
- models/deepseek-ai/DeepSeek-V4-Flash-0731 ↗
- models/Qwen/Qwen3.8-2.4T-A95B ↗
- models/Qwen/Qwen3.8-Flash-Next ↗
- models/Qwen/Qwen3.8-27B ↗
- models/deepseek-ai/DeepSeek-V4-Flash-Vision-Exp ↗
LiveBench · 多版本来源榜
LiveBench 团队，livebench.ai，2026-06-25 题集，2026-09-28 网站材料。本站进行型号映射、类别均分、两位小数舍入与范围内排序；改编成绩数据沿用 CC BY-SA 4.0。未传播题目、回答或上游代码，不代表作者背书。
已发布 47 个独立版本榜，合计覆盖 104 个型号。历史题集独立存放，不拼接成当前全科成绩。
登记题集：2026-06-25、2024-11-25、2025-05-30、2025-11-25、2025-12-23、2026-01-08。当前场景仍使用 2026-06-25。
题集及类别结构固定审核；同题集数值可更新并记录内容哈希。新增未知配置进入待审，缺任务不补零。每型号在各题集内选择全科最高配置，所有子榜沿用该配置。
CC BY-SA 4.0：本站改编类别成绩、型号关联和排序沿用相同许可，未传播题目、回答或上游程序。
各题集已发布榜单
LiveBench · 全科 · 2024-11-25 · 19 型号 →
LiveBench · 推理 · 2024-11-25 · 19 型号 →
LiveBench · 编程答题 · 2024-11-25 · 19 型号 →
LiveBench · 数学 · 2024-11-25 · 19 型号 →
LiveBench · 数据分析 · 2024-11-25 · 19 型号 →
LiveBench · 语言 · 2024-11-25 · 19 型号 →
LiveBench · 指令遵循 · 2024-11-25 · 19 型号 →
LiveBench · 全科 · 2025-05-30 · 24 型号 →
LiveBench · 推理 · 2025-05-30 · 24 型号 →
LiveBench · 编程答题 · 2025-05-30 · 24 型号 →
LiveBench · 代理编程 · 2025-05-30 · 24 型号 →
LiveBench · 数学 · 2025-05-30 · 24 型号 →
LiveBench · 数据分析 · 2025-05-30 · 24 型号 →
LiveBench · 语言 · 2025-05-30 · 24 型号 →
LiveBench · 指令遵循 · 2025-05-30 · 24 型号 →
LiveBench · 全科 · 2025-11-25 · 22 型号 →
LiveBench · 推理 · 2025-11-25 · 22 型号 →
LiveBench · 编程答题 · 2025-11-25 · 22 型号 →
LiveBench · 代理编程 · 2025-11-25 · 22 型号 →
LiveBench · 数学 · 2025-11-25 · 22 型号 →
LiveBench · 数据分析 · 2025-11-25 · 22 型号 →
LiveBench · 语言 · 2025-11-25 · 22 型号 →
LiveBench · 指令遵循 · 2025-11-25 · 22 型号 →
LiveBench · 全科 · 2025-12-23 · 22 型号 →
LiveBench · 推理 · 2025-12-23 · 22 型号 →
LiveBench · 编程答题 · 2025-12-23 · 22 型号 →
LiveBench · 代理编程 · 2025-12-23 · 22 型号 →
LiveBench · 数学 · 2025-12-23 · 22 型号 →
LiveBench · 数据分析 · 2025-12-23 · 22 型号 →
LiveBench · 语言 · 2025-12-23 · 22 型号 →
LiveBench · 指令遵循 · 2025-12-23 · 22 型号 →
LiveBench · 全科 · 2026-01-08 · 42 型号 →
LiveBench · 推理 · 2026-01-08 · 42 型号 →
LiveBench · 编程答题 · 2026-01-08 · 42 型号 →
LiveBench · 代理编程 · 2026-01-08 · 42 型号 →
LiveBench · 数学 · 2026-01-08 · 42 型号 →
LiveBench · 数据分析 · 2026-01-08 · 42 型号 →
Epoch AI · 官方指数与自有成绩
Epoch AI, Capabilities & benchmarking (2026). 本站进行型号映射、分值单位转换及子集排序；ECI 是官方拟合输出，不传播第三方原始成绩。
Epoch 自有成绩；外部 _external.csv 不准入。
当前发布 13 个榜，覆盖 123 型号。复核截止 2026-10-29。
Epoch · GPQA diamond · 179 配置 →
Epoch · MATH level 5 · 35 配置 →
Epoch · SWE-Bench verified · 26 配置 →
Epoch · OTIS Mock AIME 2024-2025 · 178 配置 →
Epoch · FrontierMath-2025-02-28-Private · 63 配置 →
Epoch · FrontierMath-Tier-4-2025-07-01-Private · 51 配置 →
Epoch · FrontierMath-Tiers-1-3-v2-Private · 91 配置 →
Epoch · FrontierMath-Tier-4-v2-Private · 56 配置 →
Epoch · SimpleQA Verified · 62 配置 →
Epoch · Chess Puzzles · 158 配置 →
Arena · 官方 Hugging Face 数据集
Arena / lmarena-ai, leaderboard-dataset. 本站筛选 overall、映射型号并重新排列收录范围名次。来源为官方主动发布数据集，不抓取 Arena 网站。
官方人类偏好 Bradley–Terry rating 与置信区间（points，不是通过率或智力）。保留子集/overall/发布时间；不接入使用 IPS 的 Agent Arena。
当前发布 4 个榜，覆盖 24 型号。复核截止 2026-10-29。
Arena · text_to_image · 人类偏好 · 4 配置 →
Arena · image_edit · 人类偏好 · 4 配置 →