How do I open the Compare Prices grid?

Two ways: click the 'Compare Prices' tab at the top of the calculator card on the home page, or navigate directly to /tools/compare-prices. The standalone page is also linked from the main navigation under 'Tools'.

Can I share my comparison with teammates?

Yes — the page URL captures the current state. Send the link in Slack and your teammate sees the same grid. Useful for procurement and architecture-review meetings.

Is the data live or cached?

Live from OpenRouter (prices) and a blended Arena AI + Artificial Analysis pipeline (quality), refreshed on a 60-minute incremental cache. So the grid is at most an hour stale.

Where do I go after the grid to project monthly cost?

Once you've picked a winner, go to /tools/api-cost-estimator and plug in the model + your expected monthly token volume. The estimator does the per-1M math against your real workload mix.

Compare Prices for ML Research: Picking Baselines and Comparators

What This Walkthrough Covers

TokenRate's new Compare Prices grid puts every model's per-token rates, context window, and quality score in a single side-by-side view. The point: stop flipping between provider pricing pages and OpenRouter tabs. You pick a provider dropdown, check the models you want, repeat for each provider, and the grid stacks every pick into one comparison table. This guide walks through using Compare Prices for picking research baselines with current pricing — a workflow that takes 90 seconds once you've done it twice. See also: filter LLM models by tier, cost, quality, Value column vs tokens-per-dollar, and how to pick an LLM by quality score and cost.

Step 1: Set Up Your Provider Dropdowns

Open /tools/compare-prices. The page loads with seven provider dropdowns: Anthropic, OpenAI, Google, Meta, DeepSeek, Mistral, xAI. For this workflow, you'll want to click through 2-3 of them and tick the models you want in the grid. The dropdowns are independent — checking a model in one doesn't affect the others, and the grid updates live as you check or uncheck.

Step 2: Read the Grid

The grid has columns for input cost per 1M, output cost per 1M, context window, and quality score. Scan vertically by column to compare like-to-like. The model ID (Anthropic's "claude-sonnet-4-7", OpenAI's "gpt-5", etc.) is what you'd paste into your SDK — useful for the copy-to-implementation step. Watch for output-to-input ratios — a 5× ratio means generation-heavy workloads cost 5× more than the input rate suggests.

Step 3: Make the Decision

Three lenses to apply: (1) is the quality score above the floor for your workload, (2) does the Value column (quality ÷ input cost) rank competitively, (3) does the output cost match your expected reply length? The grid surfaces all three in one view, which is the productivity unlock over scrolling pricing pages. For tighter quality / cost shortlisting, combine with the Filter panel on the main calculator before opening the grid.

Step 4: Operationalize

Once you've picked a winner, plug the model ID and your token volume into /tools/api-cost-estimator for a monthly bill projection. For multi-model routing decisions, also check multi-model routing with quality scores. Both the price denominator (OpenRouter) and the quality numerator (Arena AI + Artificial Analysis) refresh hourly. So the comparison you screenshot Monday morning is still trustworthy at standup Tuesday morning — but you should re-run it before a quarterly model-routing review. Try the comparison yourself at /tools/compare-prices — it's the fastest way to stack model cost, context, and quality in a single grid.