Skip to Content
API Reference

API Overview

The Ant Ling API provides a suite of large language models designed for production-grade applications. It supports a spectrum of capabilities including multi-turn conversations, deep logical reasoning (CoT), and full-modal interactions.


Prerequisites

To begin integrating with Ant Ling, ensure you have completed the following steps:

  1. Visit the API Console  to generate your unique API Token.
  2. Follow our Getting Started Guide to configure your environment variables.

Text Conversation

We recommend using Ling-2.6-1T, Ling-2.6-flash, Ling-2.5-1T and Ring-2.5-1T to perform conversation tasks, reasoning, and supports tool calls based on context.

It can be accessed via HTTP requests, OpenAI SDK, or Anthropic SDK.

Model Quick Overview

Model NameContext WindowDescription
Ling-2.6-1T256KDelivers the full chain from logical reasoning to task execution with minimal compute overhead (1/4 token cost).
Ling-2.6-flash256KExcels at code completion and problem fixing, fast document processing, and lightweight Agent interaction scenarios.
Ling-2.5-1T128KHigh-throughput decoding and leading ultra-long context processing capability.
Ling-1T128KThe first flagship non-reasoning model in the Ling series, with trillion-parameter scale.
Ring-2.5-1T128KCompared to the Ring-1T version, achieves significant improvements in key dimensions such as generation efficiency, reasoning depth, and long-cycle task execution capability.
Ring-1T128KTrained with RLVR based on the icepop method, with outstanding natural language reasoning capability, achieving SOTA results on AIME 25, CodeForces, HMMT25, LiveCodeBench, ARC-AGI-v1, and ranking first among open-source models on multiple metrics.

View Details

Was this page helpful?
Last updated on