Réponse directe

Cette page explique comment les équipes utilisent la passerelle compatible OpenAI de NextModel. Point any OpenAI-compatible coding agent at NextModel with one env block. Elle ajoute les étapes pratiques, les points de configuration et les questions fréquentes.

Env bridge

Any OpenAI-compatible agent (Claude Code env bridge, Cursor, Codex CLI) needs exactly three facts: the API base, the key, and the model id. Paste the block below into your agent config.

Agent setup
# Agent setup (any OpenAI-compatible agent: Claude Code env bridge, Cursor, Codex CLI)
export NEXTMODEL_API_KEY=$NEXTMODEL_API_KEY
export NEXTMODEL_BASE_URL=https://modelxing.com

# ~/.claude/settings.local.json
{
  "env": {
    "NEXTMODEL_API_KEY": "$NEXTMODEL_API_KEY",
    "NEXTMODEL_BASE_URL": "https://modelxing.com"
  }
}

# Model ID on this page: qwen3-max
# OpenAI-compatible endpoint: https://modelxing.com/v1/chat/completions
# Verify:
curl -s https://modelxing.com/v1/models -H "Authorization: Bearer $NEXTMODEL_API_KEY" | head -c 300

Verify

One command proves the wiring before you start a session.

Verify
curl -s https://modelxing.com/v1/models -H "Authorization: Bearer $NEXTMODEL_API_KEY" | head -c 300