Nvidia
Access 107 Nvidia models through Mastra's model router. Authentication is handled automatically using the NVIDIA_API_KEY environment variable.
Learn more in the Nvidia documentation.
.env
NVIDIA_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: "nvidia/abacusai/dracarys-llama-3.1-70b-instruct"
});
// 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);
}
备注
Mastra uses the OpenAI-compatible /chat/completions endpoint. Some provider-specific features may not be available. Check the Nvidia documentation for details.
ModelsModels的直接链接
| Model | Context | Tools | Reasoning | Image | Audio | Video | Input $/1M | Output $/1M |
|---|---|---|---|---|---|---|---|---|
nvidia/baai/bge-m3 | 8K | — | — | |||||
nvidia/black-forest-labs/flux_1-kontext-dev | 41K | — | — | |||||
nvidia/black-forest-labs/flux_1-schnell | 77 | — | — | |||||
nvidia/black-forest-labs/flux_2-klein-4b | 41K | — | — | |||||
nvidia/black-forest-labs/flux.1-dev | 4K | — | — | |||||
nvidia/deepseek-ai/deepseek-v4.1-flash | 1.0M | — | — | |||||
nvidia/google/diffusiongemma-26b-a4b-it | 250K | — | — | |||||
nvidia/google/gemma-3-12b-it | 131K | — | — | |||||
nvidia/google/gemma-3-4b-it | 131K | — | — | |||||
nvidia/google/gemma-4-31b-it | 256K | — | — | |||||
nvidia/google/google-paligemma | 128K | — | — | |||||
nvidia/meta/esm2-650m | 128K | — | — | |||||
nvidia/meta/esmfold | 128K | — | — | |||||
nvidia/meta/llama-3.2-11b-vision-instruct | 128K | — | — | |||||
nvidia/meta/llama-3.2-90b-vision-instruct | 128K | — | — | |||||
nvidia/meta/llama-guard-4-12b | 128K | — | — | |||||
nvidia/meta/muse-glimmer-30b | 131K | — | — | |||||
nvidia/mistralai/magistral-small-2506 | 33K | — | — | |||||
nvidia/mistralai/mistral-7b-instruct-v0.3 | 66K | — | — | |||||
nvidia/mistralai/mistral-medium-3-instruct | 131K | — | — | |||||
nvidia/moonshotai/kimi-k3 | 1.0M | — | — | |||||
nvidia/nvidia/active-speaker-detection | — | — | — | |||||
nvidia/nvidia/bevformer | 128K | — | — | |||||
nvidia/nvidia/cosmos-predict1-5b | — | — | — | |||||
nvidia/nvidia/cosmos-reason2-8b | 131K | — | — | |||||
nvidia/nvidia/cosmos-transfer1-7b | — | — | — | |||||
nvidia/nvidia/cosmos-transfer2_5-2b | — | — | — | |||||
nvidia/nvidia/llama-3_2-nemoretriever-300m-embed-v1 | 33K | — | — | |||||
nvidia/nvidia/llama-3.1-nemotron-70b-instruct | 128K | — | — | |||||
nvidia/nvidia/llama-3.1-nemotron-safety-guard-8b-v3 | 128K | — | — | |||||
nvidia/nvidia/llama-3.1-nemotron-ultra-253b-v1 | 128K | — | — | |||||
nvidia/nvidia/llama-nemotron-embed-vl-1b-v2 | 33K | — | — | |||||
nvidia/nvidia/llama-nemotron-rerank-vl-1b-v2 | 128K | — | — | |||||
nvidia/nvidia/magpie-tts-zeroshot | — | — | — | |||||
nvidia/nvidia/nemotron-3-nano-omni-30b-a3b-reasoning | 256K | — | — | |||||
nvidia/nvidia/nemotron-3-ultra-550b-a55b | 1.0M | $0.50 | $3 | |||||
nvidia/nvidia/nemotron-3.5-content-safety | 128K | — | — | |||||
nvidia/nvidia/nemotron-3.5-lightning-30b-a3b | 262K | — | — | |||||
nvidia/nvidia/nemotron-3.5-super-vl-120b-a12b | 1.0M | — | — | |||||
nvidia/nvidia/nemotron-voicechat | 128K | — | — | |||||
nvidia/nvidia/nv-embed-v1 | 33K | — | — | |||||
nvidia/nvidia/nv-embedcode-7b-v1 | 33K | — | — | |||||
nvidia/nvidia/rerank-qa-mistral-4b | 128K | — | — | |||||
nvidia/nvidia/riva-translate-4b-instruct-v1.1 | 128K | — | — | |||||
nvidia/nvidia/sparsedrive | 128K | — | — | |||||
nvidia/nvidia/streampetr | 128K | — | — | |||||
nvidia/nvidia/studiovoice | 128K | — | — | |||||
nvidia/nvidia/synthetic-video-detector | — | — | — | |||||
nvidia/nvidia/usdcode | 128K | — | — | |||||
nvidia/nvidia/usdvalidate | — | — | — | |||||
nvidia/openai/gpt-oss-20b | 131K | — | — | |||||
nvidia/openai/whisper-large-v3 | — | — | — | |||||
nvidia/poolside/laguna-xs-2.1 | 262K | — | — | |||||
nvidia/qwen/qwen-image | — | — | — | |||||
nvidia/qwen/qwen-image-edit | — | — | — | |||||
nvidia/stepfun-ai/step-5-preview | 1.0M | — | — | |||||
nvidia/z-ai/glm-5.3 | 1.0M | — | — | |||||
nvidia/z-ai/glm-5.3-flash | 1.0M | — | — |
Model availability, capabilities, context windows, and pricing are sourced from models.dev and may change.
Advanced configurationAdvanced configuration的直接链接
Custom headersCustom headers的直接链接
src/mastra/agents/my-agent.ts
const agent = new Agent({
id: "custom-agent",
name: "custom-agent",
model: {
url: "https://integrate.api.nvidia.com/v1",
id: "nvidia/abacusai/dracarys-llama-3.1-70b-instruct",
apiKey: process.env.NVIDIA_API_KEY,
headers: {
"X-Custom-Header": "value"
}
}
});
Dynamic model selectionDynamic model selection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
? "nvidia/z-ai/glm-5.3-flash"
: "nvidia/abacusai/dracarys-llama-3.1-70b-instruct";
}
});