Codex Fast mode omits the multiplier for GPT-6 Sol, GPT-6.1 Sol and GPT-6 Luna #1852

What happened?

Codex reports charge Fast-mode usage of gpt-6-sol, gpt-6.1-sol and gpt-6-luna at the Standard rate. fast-multiplier-overrides.json has entries for gpt-5.6-sol, gpt-5.6-terra, gpt-5.6-luna and gpt-6-astra, but none for these three, so the multiplier stays at 1.0.

Steps to reproduce

1. Use Codex with one of the three models and Fast mode (service_tier = "priority").
2. Compare ccusage codex session -j --speed standard and ccusage codex session -j --speed fast for a session that used only that model.
3. The cost is identical.

On my logs, sessions that used a single model give these fast/standard ratios: gpt-5.6-sol 2.000, gpt-5.6-terra 2.000, gpt-6-astra 2.000, gpt-6-sol 1.000, gpt-6.1-sol 1.000, gpt-6-luna 1.000.

Expected behavior

Apply the documented 2x Fast multiplier:

• https://developers.openai.com/api/docs/models/gpt-6-sol ("Fast mode is priced at 2x the applicable rates")
• https://developers.openai.com/api/docs/models/gpt-6.1-sol ("Fast mode prices are 2x Standard")
• https://developers.openai.com/api/docs/models/gpt-6-luna ("Fast mode is priced at 2x the applicable rates")

LiteLLM also lists input_cost_per_token_priority at 2x for all three.

Version

20.0.28; the entries are also missing on main.