Available Models

​

Model Overview

​

Claude Models

​

claude-sonnet-4-5

• Best for: Complex coding tasks, detailed analysis, long conversations
• Context: Up to 200K tokens
• Strengths: Code generation, debugging, technical writing, complex reasoning
• Speed: Balanced (medium)

const result = await generateText({
  model: terramind("claude-sonnet-4-5"),
  prompt: "Refactor this legacy code to use modern TypeScript patterns",
})

​

claude-sonnet-4

• Best for: General coding and analysis tasks
• Context: Up to 200K tokens
• Strengths: Code understanding, documentation, technical explanations
• Speed: Balanced (medium)

const result = await generateText({
  model: terramind("claude-sonnet-4"),
  prompt: "Explain this algorithm",
})

​

claude-opus-4-1

• Best for: Extremely complex tasks requiring deep reasoning
• Context: Up to 200K tokens
• Strengths: Research, complex problem solving, academic writing
• Speed: Slower but more thorough
• Cost: Higher

const result = await generateText({
  model: terramind("claude-opus-4-1"),
  prompt: "Design a distributed system architecture for...",
})

​

claude-haiku-4-5

• Best for: Quick responses, simple tasks, high-volume requests
• Context: Up to 200K tokens
• Strengths: Speed, efficiency, cost-effectiveness
• Speed: Very fast
• Cost: Lower

const result = await generateText({
  model: terramind("claude-haiku-4-5"),
  prompt: "Fix this syntax error",
})

​

claude-3-5-haiku

• Best for: Fast responses with good quality
• Context: Up to 200K tokens
• Speed: Fast
• Cost: Lower

const result = await generateText({
  model: terramind("claude-3-5-haiku"),
  prompt: "Summarize this code",
})

​

GPT Models

​

gpt-5

• Best for: General tasks, creative writing, reasoning
• Context: Up to 128K tokens
• Strengths: Versatility, creativity, general knowledge
• Speed: Fast

const result = await generateText({
  model: terramind("gpt-5"),
  prompt: "Write a technical blog post about WebAssembly",
})

​

gpt-5-codex

• Best for: Code generation and understanding
• Context: Up to 128K tokens
• Strengths: Code completion, bug fixing, test generation
• Speed: Fast
• Specialization: Optimized for programming tasks

const result = await generateText({
  model: terramind("gpt-5-codex"),
  prompt: "Generate unit tests for this React component",
})

​

Chinese Models

​

glm-4.6

• Best for: Chinese language tasks, bilingual content
• Strengths: Chinese understanding, translation, cultural context
• Languages: Chinese, English, and more

const result = await generateText({
  model: terramind("glm-4.6"),
  prompt: "用中文解释这个算法",
})

​

kimi-k2

• Best for: Long context understanding
• Context: Very large context window
• Strengths: Long document analysis, Chinese language

const result = await generateText({
  model: terramind("kimi-k2"),
  prompt: "分析这个长文档",
})

​

qwen3-coder

• Best for: Code generation in Chinese/English
• Strengths: Programming, Chinese code comments
• Specialization: Coding tasks

const result = await generateText({
  model: terramind("qwen3-coder"),
  prompt: "编写一个排序算法",
})

​

Specialized Models

​

grok-code

• Best for: Code analysis and generation
• Strengths: Programming tasks, technical analysis
• Specialization: Software development

const result = await generateText({
  model: terramind("grok-code"),
  prompt: "Analyze this codebase architecture",
})

​

big-pickle

• Best for: Domain-specific applications
• Use case: Custom enterprise needs

const result = await generateText({
  model: terramind("big-pickle"),
  prompt: "Process this domain-specific data",
})

​

Model Comparison

​

Choosing the Right Model

​

For Coding Tasks

• Use claude-sonnet-4-5 or claude-opus-4-1

• Use claude-haiku-4-5 or gpt-5-codex

• Use gpt-5-codex or claude-sonnet-4-5

​

For Chinese Language

• Use glm-4.6 for general tasks
• Use qwen3-coder for coding
• Use kimi-k2 for long documents

​

For Cost Optimization

• Use claude-haiku-4-5

• Use claude-sonnet-4-5 or gpt-5

• Use claude-opus-4-1

​

Model-Specific Features

​

Long Context

• Analyzing entire codebases
• Processing long documents
• Maintaining conversation history

• claude-sonnet-4-5 (200K)
• claude-opus-4-1 (200K)
• kimi-k2 (very large)

​

Streaming

const result = await streamText({
  model: terramind("claude-sonnet-4-5"),
  prompt: "Explain async programming",
})

for await (const chunk of result.textStream) {
  process.stdout.write(chunk)
}

​

Tool Calling

const result = await generateText({
  model: terramind("claude-sonnet-4-5"),
  prompt: "Calculate the area of a circle with radius 5",
  tools: {
    calculator: tool({
      description: "Perform calculations",
      parameters: z.object({ expression: z.string() }),
      execute: async ({ expression }) => eval(expression),
    }),
  },
})

​

Pricing

• Fast models (Haiku): Lower cost
• Balanced models (Sonnet, GPT-5): Medium cost
• Powerful models (Opus): Higher cost

terramind stats

​

Model Updates

terramind upgrade

npm update -g terramind

​

Next Steps