Overview

Anthropic Claude Code (slug claude) is Anthropic’s command line coding agent, run non-interactively against the test case prompt. It talks to the Anthropic API directly, so a single provider serves and bills its runs. See claude.com/claude-code.

Model IDs

Claude Code takes vendor-native Anthropic model identifiers, the same names the Anthropic API accepts, passed straight through to the CLI’s --model flag. The following are illustrative:

• claude-opus-4-8
• claude-sonnet-4-6
• claude-haiku-4-5

Invocation

The harness probes and invokes the claude binary. The CLI is installed into the run container immediately before the session, so each run picks up the most recently published version. The installer runs as the unprivileged run user and drops the binary into ~/.local/bin, which is already on PATH:

curl -fsSL https://claude.ai/install.sh | bash

Each non-interactive session runs the binary with these flags:

claude --print \
  --permission-mode bypassPermissions \
  --output-format stream-json \
  --verbose \
  --model <model> \
  <prompt>

--print runs a single non-interactive turn. --permission-mode bypassPermissions lets the agent act without approval prompts. --output-format stream-json --verbose selects the line-delimited JSON event stream the harness layer parses into events. The prompt is the final positional argument.

Authentication

Claude Code authenticates with an Anthropic API key or a Claude account subscription, preferring the subscription when its credentials are present. See Authentication for the variables and credential files each mode uses and how to lock the mode.

Pricing

Claude Code drives one provider directly and reports the exact charge for the run on its terminal result. That figure is recorded as the run’s actual cost. The comparable cost is computed from the model’s curated list price, as for every harness. See Metrics.