fix: revive cost tracking for non-DeepSeek models with an expanded pricing table

[mvanhorn]

Summary

crates/tui/src/pricing.rs pricing_for_model returns None for every model except deepseek* and Xiaomi MiMo, so the turn/session cost line, cache-savings readout, and background cost accrual are all dead when running Kimi, Qwen, GLM, MiniMax, OpenAI, Arcee, or any OpenRouter-curated model. Worse, the sidebar always renders a cost line, and format_cost_amount floors at <$0.0001, so unpriced models show cost: <$0.0001 which reads like "free" instead of "no data". The cache telemetry is already normalized for any provider (client.rs:1403-1410 maps OpenAI-style cached_tokens into Usage.prompt_cache_hit_tokens), so the math works the moment a pricing-table row exists. The issue is maintainer-authored against the active v0.8.61 milestone with a numbered implementation plan, exact file:line references, acceptance criteria, and verification commands.

Mirror the existing static curated-model match-table pattern from known_context_window_for_model (models.rs:250-279). (1) Change ModelPricing.cny to Option<CurrencyPricing> so DeepSeek/MiMo keep Some and new USD-only rows use None. (2) Extend pricing_for_model_at with a static match table for the curated catalog (kimi-k2.6, z-ai/glm-5.1, minimax/minimax-m3, qwen/qwen3.6-plus, trinity-large-thinking, gpt-5.5, etc.), sourcing prices from the OpenRouter API and omitting any row that cannot be verified. (3) Add pub fn has_pricing_for_model(model: &str) -> bool. (4) In the sidebar cost-line render site (sidebar.rs:2342-2357), when has_pricing_for_model is false AND the session total is 0.0, render cost: n/a (no pricing data for <model>) instead of <$0.0001. (5) In the cost-amount/currency path on app (app.rs), when cost_currency is CNY but the estimate has usd > 0.0 and cny == 0.0, display the USD amount instead of <¥0.0001. Keep all existing DeepSeek/MiMo behavior and tests intact.

Testing

Tests for the pricing path were added/updated alongside the change. The Rust toolchain checks (cargo fmt --all -- --check, cargo clippy --workspace --all-targets --all-features, cargo test --workspace --all-features) run in CI on this PR.

Checklist

• Added or updated tests where relevant
• Harvested/co-authored credit uses a GitHub numeric noreply address

Fixes #3066

[greptile-apps]

Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.

[gemini-code-assist]

Code Review

This pull request refactors the API cost estimation module to support USD-only models (such as those from OpenRouter) by making the CNY pricing field optional and implementing a fallback to USD display when CNY pricing is unavailable. It also adds pricing data for several new models and updates the TUI sidebar to handle unpriced models gracefully. The review feedback highlights a potential formatting bug in last_turn_cache_savings where USD savings might be incorrectly formatted with a CNY symbol, and suggests adding short aliases for the newly introduced Qwen, Gemma, and Tencent models to ensure pricing lookups succeed when short names are used.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

When self.cost_currency is Cny, but the model is USD-only (meaning estimate.cny == 0.0 and estimate.usd > 0.0), returning estimate.usd directly can cause a formatting bug. If the app's active display currency has been resolved to Cny (due to previously accrued CNY costs in the session), format_cost_amount will format this returned USD value with the CNY symbol (¥), displaying an incorrect and misleading value (e.g., ¥0.34 instead of $0.34). Matching against self.cost_display_currency(self.cost_currency) ensures that we only return USD savings when the display currency is actually resolved to USD.

        Some(match self.cost_display_currency(self.cost_currency) {
            crate::pricing::CostCurrency::Usd => estimate.usd,
            crate::pricing::CostCurrency::Cny => estimate.cny,
        })

[gemini-code-assist]

Several newly added models (Qwen, Google Gemma, and Tencent) are only matched using their full provider-prefixed names (e.g., "qwen/qwen3.6-flash"). They lack the short aliases (e.g., "qwen3.6-flash") that other models like Kimi, GLM, and Minimax have. If a user configures or runs these models using their short names, the pricing lookup will fail, resulting in no pricing data being displayed.

Suggested change

	"qwen/qwen3.6-flash" => Some(usd_only_pricing(0.1875, 0.1875, 1.125)),
	"qwen/qwen3.6-35b-a3b" => Some(usd_only_pricing(0.05, 0.15, 1.00)),
	"qwen/qwen3.6-max-preview" => Some(usd_only_pricing(1.04, 1.04, 6.24)),
	"qwen/qwen3.6-27b" => Some(usd_only_pricing(0.2885, 0.2885, 3.17)),
	"qwen/qwen3.6-plus" => Some(usd_only_pricing(0.325, 0.325, 1.95)),
	"qwen/qwen3.7-max" => Some(usd_only_pricing(0.25, 1.25, 3.75)),
	"google/gemma-4-31b-it" => Some(usd_only_pricing(0.09, 0.12, 0.35)),
	"google/gemma-4-26b-a4b-it" => Some(usd_only_pricing(0.06, 0.06, 0.33)),
	"tencent/hy3-preview" => Some(usd_only_pricing(0.021, 0.063, 0.21)),
	"qwen/qwen3.6-flash" | "qwen3.6-flash" => Some(usd_only_pricing(0.1875, 0.1875, 1.125)),
	"qwen/qwen3.6-35b-a3b" | "qwen3.6-35b-a3b" => Some(usd_only_pricing(0.05, 0.15, 1.00)),
	"qwen/qwen3.6-max-preview" | "qwen3.6-max-preview" => Some(usd_only_pricing(1.04, 1.04, 6.24)),
	"qwen/qwen3.6-27b" | "qwen3.6-27b" => Some(usd_only_pricing(0.2885, 0.2885, 3.17)),
	"qwen/qwen3.6-plus" | "qwen3.6-plus" => Some(usd_only_pricing(0.325, 0.325, 1.95)),
	"qwen/qwen3.7-max" | "qwen3.7-max" => Some(usd_only_pricing(0.25, 1.25, 3.75)),
	"google/gemma-4-31b-it" | "gemma-4-31b-it" => Some(usd_only_pricing(0.09, 0.12, 0.35)),
	"google/gemma-4-26b-a4b-it" | "gemma-4-26b-a4b-it" => Some(usd_only_pricing(0.06, 0.06, 0.33)),
	"tencent/hy3-preview" | "hy3-preview" => Some(usd_only_pricing(0.021, 0.063, 0.21)),