SynteroLinkDeveloper Docs
Verified guideLast updated: 2026-05-15Public-safe placeholders

Model Matrix and Selection Guide

Goal: help users choose the right model without guessing.

1. Quick choice

TaskRecommended modelReason
GPT defaultgpt-5.4Balanced capability and stability
GPT lightweightgpt-5.4-miniFast and suitable for simple tasks
GPT high qualitygpt-5.5Complex reasoning, long documents, code review
Claude default codingclaude-sonnet-4-6Balanced for coding, repo reading, and tool use
Claude deep reasoningclaude-opus-4-7Complex design and deep analysis
Claude fast tasksclaude-haiku-4-5Quick Q&A, summaries, classification

2. GPT series

ModelRecommended scenariosInterface
gpt-5.4Default, coding, writing, business Q&AOpenAI-compatible /v1
gpt-5.4-miniLightweight automation, cost-sensitive batch tasksOpenAI-compatible /v1
gpt-5.5High-quality output, complex reasoning, code reviewOpenAI-compatible /v1

3. Claude series

ModelRecommended scenariosInterface
claude-sonnet-4-6Default, Claude Code, coding, repo readingAnthropic-compatible
claude-opus-4-7Deep design, complex reasoningAnthropic-compatible
claude-opus-4-6High-quality text and rigorous analysisAnthropic-compatible
claude-sonnet-4-5Coding and general tasksAnthropic-compatible
claude-haiku-4-5Fast response, summary, simple classificationAnthropic-compatible

4. Selection principles

  1. If unsure, use gpt-5.4 for GPT and claude-sonnet-4-6 for Claude.
  2. For low latency or cost-sensitive jobs, choose mini / haiku.
  3. For code review, architecture, and long-context analysis, choose higher-end models.
  4. Always copy the exact model ID.

5. Base URL reference

SeriesBase URL
GPThttps://api.synterolink.com/v1
Claude Codehttps://api.synterolink.com
Claude OpenAI-compatible fallbackhttps://api.synterolink.com/v1