obul-baseten

Provide OpenAI-compatible chat completions for DeepSeek, GLM, Kimi, and GPT OSS models.

1|2|Updated Mar 2, 2026
One-click install
npx skills add https://github.com/obulai/obul-apis --skill obul-baseten
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: obul-baseten
Source: https://github.com/obulai/obul-apis/tree/main/skills/obul-baseten
Command: npx skills add https://github.com/obulai/obul-apis --skill obul-baseten

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill provides access to high-performance, OpenAI-compatible LLMs for advanced text generation and reasoning tasks, abstracting away the complexities of direct API integration.

Core Features & Use Cases

  • OpenAI-Compatible API: Seamlessly integrate with existing OpenAI-based workflows.
  • Diverse Model Selection: Access powerful models like DeepSeek, GLM, Kimi, and GPT OSS.
  • Long Context Windows: Process and generate text for large documents (up to 262k tokens).
  • Reasoning Capabilities: Utilize models specifically tuned for complex reasoning tasks.
  • Tool Calling: Enable agents to interact with external tools and functions.
  • Streaming Responses: Get real-time output for interactive applications.
  • Use Case: You need to analyze a lengthy legal document and summarize its key points, or generate creative content that requires a deep understanding of context.

Quick Start

Use the obul-baseten skill to get chat completions with the deepseek-ai/DeepSeek-V3.1 model.

Frequently Asked Questions about obul-baseten

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I get OpenAI-compatible chat completions for DeepSeek and Kimi models?

You can get OpenAI-compatible chat completions for DeepSeek and Kimi models by using the Obul proxy, which supports standard endpoints for high-performance LLM inference and streaming responses.

Can I process long context documents up to 262k tokens for reasoning tasks?

Yes, you can process long context documents up to 262k tokens for reasoning tasks. The inference endpoint supports large context windows specifically tuned for complex reasoning and analyzing lengthy documents.

Do I need an API key to use the OpenAI-compatible proxy for LLM inference?

Yes, you need an API key to use the OpenAI-compatible proxy for LLM inference. The system requires the OBUL_API_KEY environment variable to be set for authenticating your chat completion requests.

Does the Obul proxy support tool calling and streaming responses for AI agents?

Yes, the Obul proxy supports tool calling and streaming responses for AI agents. These features enable agents to interact with external functions and receive real-time output for interactive applications.

What is the best way to integrate GLM and GPT OSS models into existing OpenAI workflows?

The best way to integrate GLM and GPT OSS models into existing OpenAI workflows is via an OpenAI-compatible API endpoint, which abstracts away direct integration complexities and allows seamless interaction.

Are there limitations when using the Obul proxy for high-performance LLM inference?

Limitations when using the Obul proxy for high-performance LLM inference include being restricted to the available models like DeepSeek, GLM, Kimi, and GPT OSS, and a maximum context window of 262k tokens.