Thinking Machines

Access 38 Thinking Machines models through Mastra's model router. Authentication is handled automatically using the TINKER_API_KEY environment variable.

Learn more in the Thinking Machines documentation.

.env

TINKER_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: "thinkingmachines/Qwen/Qwen3-8B"
});

// 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);
}

note

Mastra uses the OpenAI-compatible /chat/completions endpoint. Some provider-specific features may not be available. Check the Thinking Machines documentation for details.

Models

38 available models

Model	Context	Tools	Reasoning	Image	Audio	Video	Input $/1M	Output $/1M
thinkingmachines/deepseek-ai/DeepSeek-V3.1	33K						$2	$4
thinkingmachines/deepseek-ai/DeepSeek-V4.1-Flash-1017:peft:262144	262K						$0.60	$4
thinkingmachines/deepseek-ai/DeepSeek-V4.1-Flash:peft:262144	262K						$0.60	$4
thinkingmachines/moonshotai/Kimi-K2.6	33K						$2	$5
thinkingmachines/moonshotai/Kimi-K2.6:peft:131072	131K						$2	$5
thinkingmachines/nvidia/NVIDIA-Nemotron-3-Nano-30B-A3B-BF16	66K						$0.20	$0.49
thinkingmachines/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16	66K						$0.57	$1
thinkingmachines/nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16:peft:262144	262K						$0.57	$1
thinkingmachines/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16	66K						$2	$6
thinkingmachines/nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16:peft:262144	262K						$2	$6
thinkingmachines/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16	66K						$0.18	$0.49
thinkingmachines/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16:peft:262144	262K						$0.18	$0.49
thinkingmachines/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16	66K						$0.57	$1
thinkingmachines/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16:peft:262144	262K						$0.57	$1
thinkingmachines/openai/gpt-oss-120b	33K						$0.33	$0.84
thinkingmachines/openai/gpt-oss-120b:peft:131072	131K						$0.33	$0.84
thinkingmachines/openai/gpt-oss-20b	33K						$0.18	$0.45
thinkingmachines/Qwen/Qwen3-8B	33K						$0.20	$0.60
thinkingmachines/Qwen/Qwen3.5-35B-A3B-Base	66K						$0.54	$1
thinkingmachines/Qwen/Qwen3.5-397B-A17B	66K						$2	$8
thinkingmachines/Qwen/Qwen3.5-397B-A17B:peft:262144	262K						$2	$8
thinkingmachines/Qwen/Qwen3.5-4B	66K						$0.33	$1
thinkingmachines/Qwen/Qwen3.5-4B:peft:262144	262K						$0.33	$1
thinkingmachines/Qwen/Qwen3.5-9B	66K						$0.30	$0.81
thinkingmachines/Qwen/Qwen3.5-9B-Base	66K						$0.30	$0.81
thinkingmachines/Qwen/Qwen3.6-27B	66K						$2	$6
thinkingmachines/Qwen/Qwen3.6-35B-A3B	66K						$0.45	$1
thinkingmachines/Qwen/Qwen3.6-35B-A3B:peft:262144	262K						$0.45	$1
thinkingmachines/Qwen/Qwen3.8-27B	66K						$0.70	$6
thinkingmachines/Qwen/Qwen3.8-27B:peft:262144	262K						$0.70	$6
thinkingmachines/thinkingmachines/Inkling	66K						$2	$5
thinkingmachines/thinkingmachines/Inkling-Small	66K						$0.58	$1
thinkingmachines/thinkingmachines/Inkling-Small:peft:262144	262K						$0.58	$1
thinkingmachines/thinkingmachines/Inkling-Small:peft:262144:sampling-nvfp4	262K						$0.30	$1
thinkingmachines/thinkingmachines/Inkling:peft:262144	262K						$2	$5
thinkingmachines/thinkingmachines/Inkling:peft:262144:sampling-nvfp4	262K						$1	$4
thinkingmachines/zai-org/GLM-5.3-Flash:peft:262144	262K						$0.84	$2
thinkingmachines/zai-org/GLM-5.3:peft:262144	262K						$4	$12

Model availability, capabilities, context windows, and pricing are sourced from models.dev and may change.

Advanced configuration

Custom headers

src/mastra/agents/my-agent.ts

import { Agent } from "@mastra/core/agent";

const agent = new Agent({
  id: "custom-agent",
  name: "custom-agent",
  model: {
    url: "https://tinker.thinkingmachines.dev/services/tinker-prod/anthropic/api/v1",
    id: "thinkingmachines/Qwen/Qwen3-8B",
    apiKey: process.env.TINKER_API_KEY,
    headers: {
      "X-Custom-Header": "value"
    }
  }
});

Dynamic model selection

src/mastra/agents/my-agent.ts

import { Agent } from "@mastra/core/agent";

const agent = new Agent({
  id: "dynamic-agent",
  name: "Dynamic Agent",
  model: ({ requestContext }) => {
    const useAdvanced = requestContext.task === "complex";
    return useAdvanced
      ? "thinkingmachines/zai-org/GLM-5.3:peft:262144"
      : "thinkingmachines/Qwen/Qwen3-8B";
  }
});