跳到主要内容

Pioneer 标志Pioneer

通过 Mastra 的模型路由器访问 102 个 Pioneer 模型。系统会自动使用 PIONEER_API_KEY 环境变量完成身份验证。

更多信息请参阅Pioneer 文档。

.env
PIONEER_API_KEY=your-api-key
src/mastra/agents/my-agent.ts
import { Agent } from '@mastra/core/agent'

const agent = new Agent({
  id: 'my-agent',
  name: 'My Agent',
  instructions: 'You are a helpful assistant',
  model: 'pioneer/HuggingFaceTB/SmolLM3-3B-Base',
})

// Generate a response
const response = await agent.generate('Hello!')

// Stream a response
const stream = await agent.stream('Tell me a story')
for await (const chunk of stream) {
  console.log(chunk)
}
信息

Mastra 使用兼容 OpenAI 的 /chat/completions endpoint。部分 Provider 特有功能可能不可用，详情请参阅Pioneer 文档。

模型
模型的直接链接

ModelContextToolsReasoningImageAudioVideoInput $/1MOutput $/1M
pioneer/claude-3-7-sonnet-latest1.0M$3$15
pioneer/claude-fable-51.0M$11$55
pioneer/claude-haiku-4-5200K$1$5
pioneer/claude-opus-4-1200K$15$75
pioneer/claude-opus-4-5200K$5$25
pioneer/claude-opus-4-61.0M$5$25
pioneer/claude-opus-4-71.0M$5$25
pioneer/claude-opus-4-81.0M$5$25
pioneer/claude-opus-51.0M$5$25
pioneer/claude-sonnet-4-51.0M$3$15
pioneer/claude-sonnet-4-61.0M$3$15
pioneer/claude-sonnet-51.0M$2$10
pioneer/deepseek-ai/DeepSeek-V3164K$0.27$1
pioneer/deepseek-ai/DeepSeek-V3.1164K$0.56$2
pioneer/deepseek-ai/DeepSeek-V4-Flash1.0M$0.10$0.20
pioneer/deepseek-ai/DeepSeek-V4-Pro1.0M$0.43$0.87
pioneer/devstral-2256K$0.40$2
pioneer/fastino/gliguard-LLMGuardrails-300M8K$0.15$0.15
pioneer/fastino/gliner2-base-v18K$0.15$0.15
pioneer/fastino/gliner2-large-v18K$0.15$0.15
pioneer/fastino/gliner2-multi-large-v18K$0.15$0.15
pioneer/fastino/gliner2-multi-v18K$0.15$0.15
pioneer/fastino/gliner2-privacy-filter-PII-multi8K$0.15$0.15
pioneer/gemini-3-flash1.0M$0.50$3
pioneer/gemini-3.1-flash-lite1.0M$0.25$2
pioneer/gemini-3.1-pro1.0M$2$12
pioneer/gemini-3.5-flash1.0M$2$9
pioneer/gemini-3.5-flash-lite1.0M$0.30$3
pioneer/gemini-3.6-flash1.0M$2$8
pioneer/google/diffusiongemma-26B-A4B-it262K$0.50$0.50
pioneer/google/gemma-3-4b-pt33K$0.15$0.15
pioneer/google/gemma-4-12B-it33K$0.25$0.25
pioneer/google/gemma-4-31B-it33K$0.50$0.50
pioneer/google/gemma-4-E2B-it33K$0.10$0.10
pioneer/google/gemma-4-E4B-it33K$0.20$0.20
pioneer/gpt-4.11.0M$2$8
pioneer/gpt-4.1-mini1.0M$0.40$2
pioneer/gpt-4.1-nano1.0M$0.10$0.40
pioneer/gpt-4o128K$3$10
pioneer/gpt-4o-mini128K$0.15$0.60
pioneer/gpt-5-mini400K$0.25$2
pioneer/gpt-5-nano400K$0.05$0.40
pioneer/gpt-5.1400K$1$10
pioneer/gpt-5.3-codex400K$2$14
pioneer/gpt-5.41.1M$3$15
pioneer/gpt-5.4-mini400K$0.75$5
pioneer/gpt-5.4-nano1.0M$0.20$1
pioneer/gpt-5.51.1M$5$30
pioneer/gpt-5.6-luna1.1M$1$6
pioneer/gpt-5.6-sol1.1M$5$30
pioneer/gpt-5.6-terra1.1M$3$15
pioneer/grok-4.5500K$2$6
pioneer/HuggingFaceTB/SmolLM3-3B-Base33K$0.15$0.15
pioneer/LiquidAI/LFM2-24B-A2B33K$0.03$0.12
pioneer/meta-llama/Llama-3.1-8B-Instruct131K$0.20$0.20
pioneer/meta-llama/Llama-3.2-1B131K$0.10$0.10
pioneer/meta-llama/Llama-3.2-1B-Instruct131K$0.10$0.20
pioneer/meta-llama/Llama-3.2-3B131K$0.10$0.10
pioneer/meta-llama/Llama-3.2-3B-Instruct131K$0.10$0.34
pioneer/meta-llama/Llama-3.3-70B-Instruct131K$0.90$0.90
pioneer/meta/muse-spark-1.11.0M$1$4
pioneer/MiniMaxAI/MiniMax-M2.7205K$0.28$1
pioneer/MiniMaxAI/MiniMax-M31.0M$0.30$1
pioneer/mistral-large-3256K$0.50$2
pioneer/mistral-medium-3.5262K$2$8
pioneer/mistralai/Codestral-22B-v0.1128K$0.30$0.90
pioneer/mistralai/Magistral-Small-2506128K$0.50$2
pioneer/mistralai/Ministral-8B-Instruct-2410128K$0.15$0.15
pioneer/mistralai/Mistral-7B-Instruct-v0.333K$0.20$0.20
pioneer/mistralai/Mistral-Nemo-Instruct-2407131K$0.02$0.03
pioneer/mistralai/Mistral-Small-4-119B-2603262K$0.15$0.60
pioneer/mistralai/Pixtral-12B-2409128K$0.15$0.15
pioneer/moonshotai/Kimi-K2.6262K$0.95$4
pioneer/moonshotai/Kimi-K2.7-Code262K$0.95$4
pioneer/moonshotai/Kimi-K31.0M$3$15
pioneer/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16262K$0.05$0.20
pioneer/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP81.0M$0.09$0.45
pioneer/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF161.0M$0.50$3
pioneer/openai/gpt-oss-120b131K$0.15$0.60
pioneer/openai/gpt-oss-20b131K$0.07$0.30
pioneer/pioneer/auto1.0M——
pioneer/poolside/laguna-s-2.11.0M$0.10$0.20
pioneer/Qwen/Qwen2.5-Coder-0.5B33K$0.10$0.10
pioneer/Qwen/Qwen3-1.7B-Base33K$0.10$0.10
pioneer/Qwen/Qwen3-235B-A22B-Instruct-2507262K$1$1
pioneer/Qwen/Qwen3-32B131K$0.90$0.90
pioneer/Qwen/Qwen3-4B-Base33K$0.15$0.15
pioneer/Qwen/Qwen3-4B-Instruct-2507262K$0.20$0.20
pioneer/Qwen/Qwen3-8B131K$0.20$0.20
pioneer/Qwen/Qwen3.5-9B33K$0.30$0.30
pioneer/Qwen/Qwen3.6-27B33K$0.60$0.60
pioneer/Qwen/Qwen3.6-35B-A3B262K$0.14$1
pioneer/qwen3.6-flash1.0M$0.19$1
pioneer/qwen3.6-max-preview262K$1$6
pioneer/qwen3.6-plus1.0M$0.33$2
pioneer/qwen3.7-max1.0M$1$4
pioneer/qwen3.7-plus1.0M$0.32$1
pioneer/sakana/fugu-ultra1.0M$5$30
pioneer/XiaomiMiMo/MiMo-V2.51.1M$0.14$0.28
pioneer/XiaomiMiMo/MiMo-V2.5-Pro1.1M$0.43$0.87
pioneer/zai-org/GLM-5.1203K$0.98$3
pioneer/zai-org/GLM-5.21.0M$1$4
102 available models

高级配置
高级配置的直接链接

自定义请求头
自定义请求头的直接链接

src/mastra/agents/my-agent.ts
const agent = new Agent({
  id: 'custom-agent',
  name: 'custom-agent',
  model: {
    url: 'https://api.pioneer.ai/v1',
    id: 'pioneer/HuggingFaceTB/SmolLM3-3B-Base',
    apiKey: process.env.PIONEER_API_KEY,
    headers: {
      'X-Custom-Header': 'value',
    },
  },
})

动态选择模型
动态选择模型的直接链接

src/mastra/agents/my-agent.ts
const agent = new Agent({
  id: 'dynamic-agent',
  name: 'Dynamic Agent',
  model: ({ requestContext }) => {
    const useAdvanced = requestContext.task === 'complex'
    return useAdvanced ? 'pioneer/zai-org/GLM-5.2' : 'pioneer/HuggingFaceTB/SmolLM3-3B-Base'
  },
})
On this page