Grok Build - Kyma API

Overview

Created by xAI, this model specializes in software development tasks. It handles tool calls natively, supports structured outputs, and includes built-in reasoning capabilities. It operates exclusively with text inputs and outputs, with a 256,000-token context window and a 32,768-token maximum output limit. On Kyma, requests route through an OpenAI-compatible endpoint using a single API key. The platform provides automatic failover if a serving path degrades, and every response returns the exact token cost in usage.cost alongside the X-Kyma-Model header. Prompt caching is supported, billing repeated prefixes at this model’s cached input rate. It does not process images, audio, or other non-text modalities. While optimized for speed and tool use, its output length caps at 32,768 tokens, which may require chunking for very long codebase generation or extensive documentation tasks.

Specs

Field	Value
Model ID	`grok-build`
Best for	Agentic coding, debugging, fast tool use
Context window	256,000
Max output tokens	33K
Input modalities	Text
Output modalities	Text
Tool calling	Yes
Structured outputs	Yes
Prompt caching	Yes
Speed	Fast
Cost band	Balanced
Release stage	Preview

Pricing

	Per 1M tokens
Input	$1.389
Output	$2.777

Above a long prompt this model costs more per token, because the infrastructure serving it charges more there. The threshold is on the prompt, so you know which rate applies before you send.

Prompt length	Input	Output
Standard	$1.389	$2.777
Above 200,000 tokens	$2.70	$5.40

Use this when

Agentic Software Development — Automates multi-step coding tasks by chaining tool calls and reasoning loops.
Rapid Debugging Workflows — Identifies and patches code errors quickly using native tool integration.
Structured API Responses — Generates valid JSON or schema-compliant outputs for downstream system integration.
Fast Chat Interactions — Delivers low-latency conversational responses for developer-focused assistants.

Not ideal for

Do not use this model for tasks requiring image analysis, audio processing, or outputs longer than 32,768 tokens.

Pick something else when

You need to analyze images or process multimodal data: use gemini-3.5-flash or qwen-3.7-plus.
You require output sequences exceeding 32,768 tokens: use deepseek-v3 or qwen-3.7-max.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="kyma-...")

response = client.chat.completions.create(
    model="grok-build",
    messages=[{"role": "user", "content": "..."}],
    tools=[...],  # function calling supported
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=chat&tools=true&quality_tier=strong"

FAQ

Does Grok Build support image inputs? No, it only accepts and generates text. Use a multimodal model if your workflow requires vision. How does prompt caching affect my costs? Repeated prompt prefixes are billed at this model’s cached input rate, reducing costs for stable system prompts or long conversation histories. What happens if the serving endpoint fails during a request? Kyma automatically reroutes the request to a healthy serving path, and the response includes the exact cost in usage.cost and the active model in the X-Kyma-Model header.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ