Text Models
Text models use POST /v1/responses or compatible POST /v1/chat/completions, depending on the model. They share the same HiAPI API key as media models, but do not use the asynchronous /v1/tasks request shape.
| Model | Best for | Endpoint |
|---|---|---|
| Claude Opus 5.5 | Long-running agentic coding, knowledge work, and five effort levels | POST /v1/chat/completions · thinking always on |
| Claude Opus 5.5 (AWS) | Same model served via the AWS route with independent capacity and live rates | POST /v1/chat/completions · thinking always on |
| Claude Haiku 5.5 | Fast text and coding workflows with native Chat Completions reasoning | POST /v1/chat/completions · reasoning_effort low/medium/high/xhigh |
| Grok 4.6 | Text reasoning, streaming, JSON objects, and function tools | POST /v1/chat/completions |
| Grok 4.7 | Text reasoning, streaming, JSON objects, and function tools | POST /v1/chat/completions |
| GPT-6 Astra | Reasoning, code review, function calls, and structured output | POST /v1/responses · streaming only |
| GPT-6.1 Sol | Reasoning, coding, analysis, and streamed text generation | POST /v1/responses · streaming only |
| GPT-6.1 Terra | Fast visual reasoning, computer use, and everyday office tasks | POST /v1/responses + Chat Completions |
| GPT-6 Luna | Summaries, information extraction, and high-volume quick Q&A | POST /v1/responses + Chat Completions |
| Claude Sonnet 4.6 | Text conversations, coding assistance, and tool workflows | POST /v1/chat/completions · streaming supported |
| Kimi K3.1 Pro | Coding, agent tool workflows, and long-document reasoning | POST /v1/chat/completions · always-on reasoning |
| Kimi K3.1 Pro Highspeed | Faster Kimi K3.1 Pro variant for latency-sensitive coding, agent, and long-document work | POST /v1/chat/completions · always-on reasoning |
| Kimi K3 | Long-context reasoning, coding, structured output, and tool workflows | POST /v1/chat/completions · always-on reasoning |
| DeepSeek V4.1 Flash | Text and image input, reasoning, and streaming | POST /v1/responses + Chat |
| DeepSeek V4 Pro | Complex reasoning, coding Agents, and long-context work | POST /v1/responses + Chat compatible |
| DeepSeek V4 Flash Vision Exp | Image understanding, visual questions, function tools | POST /v1/responses + Chat compatible |
| DeepSeek V4 Flash | High-throughput generation, reasoning, coding, and Agent workflows | POST /v1/chat/completions |
| GLM-5.3 | Text generation through the OpenAI-compatible Chat Completions API | POST /v1/chat/completions |
Start with the Responses guide or Chat Completions guide, then compare the AI Text APIs collection.