实时同步 OpenRouter
OpenRouter 免费模型大全
汇总 OpenRouter 全部 29 个免费可用 AI 模型，无需付费、无需信用卡，注册即可调用。
✓ 每小时同步✓ 包含上下文长度✓ 含使用示例
⚡ 快速开始
from openai import OpenAI
client = OpenAI(
api_key="your-openrouter-api-key",
base_url="https://openrouter.ai/api/v1"
)
response = client.chat.completions.create(
model="meta-llama/llama-3.3-70b-instruct:free", # 任意下方 Model ID
messages=[{"role": "user", "content": "你好"}]
)
print(response.choices[0].message.content)API Key 在 openrouter.ai/keys 免费获取，无需信用卡。
全部免费模型 （29 个）
| 模型名称 | 提供商 | 上下文 | 能力 | Model ID |
|---|---|---|---|---|
Arcee AI: Trinity Large Preview Trinity-Large-Preview is a frontier-scale open-weight language model from Arcee, built as a 400B-par… | Arcee AI | 131K | 免费 | arcee-ai/trinity-large-preview:free |
Elephant Elephant Alpha is a 100B-parameter text model focused on intelligence efficiency, delivering strong … | OpenRouter | 262K | 免费 | openrouter/elephant-alpha |
Free Models Router The simplest way to get free inference. openrouter/free is a router that selects free models at rand… | OpenRouter | 200K | 免费视觉自动路由 | openrouter/free |
Google: Gemma 3 12B Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles cont… | 33K | 免费视觉 | google/gemma-3-12b-it:free | |
Google: Gemma 3 27B Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles cont… | 131K | 免费视觉 | google/gemma-3-27b-it:free | |
Google: Gemma 3 4B Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles cont… | 33K | 免费视觉 | google/gemma-3-4b-it:free | |
Google: Gemma 3n 2B Gemma 3n E2B IT is a multimodal, instruction-tuned model developed by Google DeepMind, designed to o… | 8K | 免费 | google/gemma-3n-e2b-it:free | |
Google: Gemma 3n 4B Gemma 3n E4B-it is optimized for efficient execution on mobile and low-resource devices, such as pho… | 8K | 免费 | google/gemma-3n-e4b-it:free | |
Google: Gemma 4 26B A4B Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Desp… | 262K | 免费视觉 | google/gemma-4-26b-a4b-it:free | |
Google: Gemma 4 31B Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image inp… | 262K | 免费视觉 | google/gemma-4-31b-it:free | |
Google: Lyria 3 Clip Preview 30 second duration clips are priced at $0.04 per clip. Lyria 3 is Google's family of music generatio… | 1M | 免费视觉 | google/lyria-3-clip-preview | |
Google: Lyria 3 Pro Preview Full-length songs are priced at $0.08 per song. Lyria 3 is Google's family of music generation model… | 1M | 免费视觉 | google/lyria-3-pro-preview | |
LiquidAI: LFM2.5-1.2B-Instruct LFM2.5-1.2B-Instruct is a compact, high-performance instruction-tuned model built for fast on-device… | LiquidAI | 33K | 免费 | liquid/lfm-2.5-1.2b-instruct:free |
LiquidAI: LFM2.5-1.2B-Thinking LFM2.5-1.2B-Thinking is a lightweight reasoning-focused model optimized for agentic tasks, data extr… | LiquidAI | 33K | 免费推理 | liquid/lfm-2.5-1.2b-thinking:free |
Meta: Llama 3.2 3B Instruct Llama 3.2 3B is a 3-billion-parameter multilingual large language model, optimized for advanced natu… | Meta | 131K | 免费 | meta-llama/llama-3.2-3b-instruct:free |
Meta: Llama 3.3 70B Instruct The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned gen… | Meta | 66K | 免费 | meta-llama/llama-3.3-70b-instruct:free |
Meta: Llama Guard 4 12B Llama Guard 4 is a Llama 4 Scout-derived multimodal pretrained model, fine-tuned for content safety … | Meta | 164K | 免费视觉 | meta-llama/llama-guard-4-12b:free |
MiniMax: MiniMax M2.5 MiniMax-M2.5 is a SOTA large language model designed for real-world productivity. Trained in a diver… | MiniMax | 197K | 免费 | minimax/minimax-m2.5:free |
Nous: Hermes 3 405B Instruct Hermes 3 is a generalist language model with many improvements over Hermes 2, including advanced age… | Nous Research | 131K | 免费 | nousresearch/hermes-3-llama-3.1-405b:free |
NVIDIA: Nemotron 3 Nano 30B A3B NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and acc… | NVIDIA | 256K | 免费 | nvidia/nemotron-3-nano-30b-a3b:free |
NVIDIA: Nemotron 3 Super NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters fo… | NVIDIA | 262K | 免费 | nvidia/nemotron-3-super-120b-a12b:free |
NVIDIA: Nemotron Nano 12B 2 VL NVIDIA Nemotron Nano 2 VL is a 12-billion-parameter open multimodal reasoning model designed for vid… | NVIDIA | 128K | 免费视觉 | nvidia/nemotron-nano-12b-v2-vl:free |
NVIDIA: Nemotron Nano 9B V2 NVIDIA-Nemotron-Nano-9B-v2 is a large language model (LLM) trained from scratch by NVIDIA, and desig… | NVIDIA | 128K | 免费 | nvidia/nemotron-nano-9b-v2:free |
OpenAI: gpt-oss-120b gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI d… | OpenAI | 131K | 免费 | openai/gpt-oss-120b:free |
OpenAI: gpt-oss-20b gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. I… | OpenAI | 131K | 免费 | openai/gpt-oss-20b:free |
Qwen: Qwen3 Coder 480B A35B Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the … | Qwen / Alibaba | 262K | 免费代码 | qwen/qwen3-coder:free |
Qwen: Qwen3 Next 80B A3B Instruct Qwen3-Next-80B-A3B-Instruct is an instruction-tuned chat model in the Qwen3-Next series optimized fo… | Qwen / Alibaba | 262K | 免费 | qwen/qwen3-next-80b-a3b-instruct:free |
Venice: Uncensored Venice Uncensored Dolphin Mistral 24B Venice Edition is a fine-tuned variant of Mistral-Small-24B-In… | Venice | 33K | 免费 | cognitivecomputations/dolphin-mistral-24b-venice-edition:free |
Z.ai: GLM 4.5 Air GLM-4.5-Air is the lightweight variant of our latest flagship model family, also purpose-built for a… | Z.ai | 131K | 免费 | z-ai/glm-4.5-air:free |
什么是 OpenRouter 免费模型？
OpenRouter 是全球领先的 AI 模型聚合 API 平台，支持一个 API Key 调用 200+ 个大模型。 其中部分模型提供完全免费的调用额度（model id 以 :free 结尾）， 包括 Meta Llama 3.3 70B、Google Gemma 3 27B、Qwen3 等顶级开源模型。 免费版与付费版使用相同的模型权重，能力无差别，适合个人开发者学习、测试和低频场景。
免费模型使用限制
- • 每日请求次数有限额（通常 20–200 次/天）
- • 高峰期响应速度比付费慢，可能排队
- • 不保证 SLA 和 99.9% 可用性
- • 适合测试、原型开发，不建议生产重度依赖
- • 超出免费限额可切换同名付费模型（去掉
:free）
推荐免费模型选择
- 🏆 综合最强：Llama 3.3 70B / Qwen3 Coder 480B
- ⚡ 速度最快：Gemma 3 4B / Qwen3 4B
- 🧠 推理代码：Qwen3 Coder / NVIDIA Nemotron
- 👁️ 多模态视觉：Gemma 3 27B / NVIDIA Nano VL
- 💬 中文最佳：Qwen3 / GLM 4.5 Air / MiniMax M2.5
国内如何使用 OpenRouter 免费 API？
OpenRouter 在国内需要代理访问。推荐通过本站收录的 AI API 中转站， 部分中转站已支持 OpenRouter 格式，可国内直连无需翻墙。 也可通过 VPN 直接访问 openrouter.ai/api/v1。
常见问题
OpenRouter 免费模型怎么用？
注册 OpenRouter 账号获取 API Key（openrouter.ai/keys），将 base_url 设为 https://openrouter.ai/api/v1，model 填写对应的 model id（如 meta-llama/llama-3.3-70b-instruct:free），即可免费调用。完全兼容 OpenAI SDK，只需替换 base_url 和 api_key 两个参数。
OpenRouter 免费模型每天能调用多少次？
大多数免费模型每天限额 20–200 次请求，具体视模型而定。超出限额会返回 429 错误，可等到次日重置，或切换到同名付费版本（去掉 :free 后缀）。
免费版和付费版模型有什么区别？
使用完全相同的模型权重，输出质量无差别。区别在于：付费版请求优先级更高、无配额限制、高峰期响应更快、有 SLA 保障。免费版适合测试和低频场景，生产环境建议使用付费版。
哪个 OpenRouter 免费模型最好用？
综合能力推荐 Meta Llama 3.3 70B（通用）、Qwen3 Coder 480B（代码）、Google Gemma 3 27B（多语言+视觉）、Qwen3/GLM 4.5 Air（中文）。具体选择取决于你的使用场景。
国内能直接用 OpenRouter 免费 API 吗？
OpenRouter 官方站点在国内需要代理访问。可以通过本站收录的 AI API 中转站（支持 OpenRouter 格式的）国内直连调用，或自行通过 VPN 直连 openrouter.ai/api/v1。