Claude Code and Codex

Add SolCard Inference to Claude Code or Codex with one sentence.

Paste this into Claude Code or Codex:

Add SolCard Inference to my models: solcardinference.ai/install

The agent reads that page, asks for your key (make one at app.solcardinference.ai), saves it in its settings and checks it. To skip the question, add your key: Save my key sk-si-... in my settings. The app’s Quickstart copies the sentence with a new key filled in.

Then restart the tool:

Claude Code: exit and run claude again. /model lists the SolCard Inference models, Large by default.
Codex: quit, run codex app-server daemon restart, and start Codex again. In the Codex app, quit and reopen it. The model picker lists the SolCard Inference models.

To switch back, ask the agent to “remove SolCard Inference from my models”.

What the install changes

solcardinference.ai/install is the exact configuration, written for agents:

Claude Code: ~/.claude/settings.json points ANTHROPIC_BASE_URL at https://api.solcardinference.ai with your key, lists the SolCard models in /model, and maps the Opus, Sonnet, Haiku and Fable tiers to SolCard models, so subagents and background tasks stay on SolCard.
Codex: ~/.codex/config.toml gets a solcard-inference provider (Responses API, the model catalog, and your key read from ~/.codex/solcard-inference.key) and makes it the default.