Roder
Changelog
9 July 2026roder/v0.1.11

Codex GPT-5.6 models and Ultra mode land in Roder

Roder 0.1.11 updates the Codex provider catalog with GPT-5.6 Sol, Terra, and Luna, makes Sol the default Codex model, and adds Ultra as a first-class effort for bounded multi-agent work.

What changed

Roder 0.1.11 refreshes the Codex OAuth model catalog. The Codex provider now lists GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, and GPT-5.4 alongside the existing GPT-5.5, GPT-5.4 Mini, and GPT-5.3 Codex Spark entries. New Codex sessions default to codex/gpt-5.6-sol.

The release also adds ultra as a Roder reasoning effort for GPT-5.6 Sol and GPT-5.6 Terra. Ultra stays visible in Roder’s provider and selection surfaces, while the OpenAI Responses transport sends the provider’s max reasoning effort on the wire. GPT-5.6 Luna remains a lighter option with low through max reasoning choices, but does not expose Ultra.

Why it matters

  • Codex users can select the current GPT-5.6 coding models directly from Roder instead of relying on stale model ids or custom catalog overrides.
  • The default Codex path now points at Sol, so a new Codex-backed run starts on the strongest current Roder catalog entry.
  • Terra and Luna give teams clearer tradeoffs: Terra for everyday agentic coding with the full Ultra option, Luna for faster or cheaper work that still has the new 372k-token context window.
  • Ultra separates Roder’s orchestration choice from the provider wire value. People can see that they chose Ultra in Roder, while the provider receives the maximum supported reasoning setting.
  • Ultra-enabled Sol and Terra turns can use Roder’s bounded in-process delegation policy. Lower efforts keep teammate spawning explicit-request-only, which makes the cost and concurrency change deliberate.

Using it

Sign in to the Codex provider as before:

roder auth login codex

Then select the model through the TUI provider picker, the app-server provider methods, or environment defaults:

RODER_PROVIDER=codex RODER_MODEL=gpt-5.6-sol roder
RODER_PROVIDER=codex RODER_MODEL=gpt-5.6-terra RODER_REASONING=ultra roder
RODER_PROVIDER=codex RODER_MODEL=gpt-5.6-luna RODER_REASONING=max roder

Use Ultra only when the task benefits from proactive, bounded delegation and the signed-in Codex account can support the extra work. For ordinary focused edits, lower reasoning efforts remain available and keep subagents explicit.