Add GPT-6 Astra pricing

[btsouth]

Fixes #413

GPT-6 Astra tokens scan fine from Codex logs but were missing from the built-in pricing table, so the Charts page showed $0.00 with "0% of tokens priced". This adds the model at OpenAI's published standard rates: $10 input / $1 cached input / $50 output per 1M tokens.

Details:

• Codex usage of Astra does not incur the >272K long-context multipliers and Codex does not charge cache writes, so the entry has no long_context rates (short-context rates apply at every context size).
• Bare gpt-6 normalizes to gpt-6-astra, matching the existing gpt-5.6 to gpt-5.6-sol pattern. Date-suffixed and -codex variants already resolve through existing logic.
• Added a regression test covering variant normalization and the published rates.

Commands run: cargo test --manifest-path rust/Cargo.toml (1186 passed), cargo test --manifest-path apps/desktop-tauri/src-tauri/Cargo.toml (617 passed), cargo fmt --all. Clippy is clean for the touched files; the two existing -D warnings failures are pre-existing on clean main (Windows-gated code that is dead on Linux).

Note

Add gpt-6-astra pricing and alias gpt-6 to it in CODEX_PRICING

• Adds a gpt-6-astra entry to CODEX_PRICING at $10/M input, $50/M output, and $1/M cached-input tokens, with no long-context rate override.
• CostUsagePricing::normalize_codex_model now maps the exact name gpt-6 to the canonical key gpt-6-astra, so the alias and dated/Codex-suffixed Astra variants resolve to the new rates.
• Adds tests covering canonical, dated, Codex-suffixed, and gpt-6 alias normalization, plus cached and long-context cost assertions.
• Behavioral Change: gpt-6 and Astra variants now resolve to gpt-6-astra pricing in cost_pricing.rs; review normalize_codex_model and the gpt-6-astra table entry to confirm rates and alias handling.

^{Macroscope summarized 79c1041.}

Summary by CodeRabbit

• New Features

  • Added pricing support for the GPT-6 Astra model, including standard input, output, and cached-input rates.
  • Added recognition of the gpt-6 model alias and related model identifiers.

• Bug Fixes

  • Ensured extended input lengths use the correct standard pricing when no long-context rates apply.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Team

Run ID: d349a670-1c87-4836-979c-cbf29b8df2f4

📥 Commits

Reviewing files that changed from the base of the PR and between 45a2a4f and 79c1041.

📒 Files selected for processing (2)

• rust/src/core/cost_pricing.rs
• rust/src/core/cost_pricing_tests.rs

Included review availability: Your plan provides up to 1 included review per hour; 0 remain after this review.

---

📝 Walkthrough

Walkthrough

Adds GPT-6 Astra pricing and maps the gpt-6 alias to gpt-6-astra. Tests cover dated aliases, Codex aliases, standard rates, cached-input billing, and 300K-token input pricing.

Changes

GPT-6 Astra pricing

Layer / File(s)	Summary
Pricing, normalization, and validation rust/src/core/cost_pricing.rs, rust/src/core/cost_pricing_tests.rs	Adds standard GPT-6 Astra input, output, and cache-read rates. Maps gpt-6 to gpt-6-astra. Tests alias normalization and pricing without long-context multipliers.

Estimated code review effort: 2 (Simple) | ~10 minutes

Merge Risk: ⚪ Minimal · up to 79c10

GPT-6 Astra usage will now resolve to published pricing rather than appearing as unpriced usage, with tested aliases and token-rate behavior. No current merge-blocking risk remains.

Suggested reviewers: finesssee

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly and concisely describes the main change: adding GPT-6 Astra pricing.
Linked Issues check	✅ Passed	The changes address issue #413 by adding GPT-6 Astra pricing, model normalization, and regression coverage so Astra usage can be recognized and displayed correctly.
Out of Scope Changes check	✅ Passed	All changes are related to GPT-6 Astra pricing, normalization, and tests. No unrelated code changes are identified.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%. Docstring coverage is scoped to functions touched by this diff. Analyzed 4 functions across 2 files.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch fix/gpt-6-astra-pricing

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[cloudflare-workers-and-pages]

Deploying with    Cloudflare Workers

The latest updates on your project. Learn more about integrating Git with Workers.

Status	Name	Latest Commit	Preview URL	Updated (UTC)
✅ Deployment successful! View logs	ceiling	79c1041	Commit Preview URL Branch Preview URL	Sep 07 2026, 02:42 PM