What will AI cost you to run?
Answer two questions and get a recommendation in plain dollars, with a cheaper and a more capable option beside it.
Prices from each provider’s official pricing page · checked recently
Best value for you
…
… / month
Every model below works with PC Investigator. Choose the provider in Settings → AI & Models, paste your key and the model ID, then press Test connection.
Every model at your usage
Cheapest first · select a model for details- Loading current prices…
How we calculate this
- Monthly cost is your estimated input tokens times each model’s input price, plus output tokens times its output price, using published list prices.
- For PC Investigator we assume about 80,000 tokens in and 4,000 out per investigation (two AI passes), and about 20,000 in and 600 out per chat message, based on the Hub’s own limits.
- For chat and writing we assume 0.3, 1 or 3 million tokens per person a month (light, regular, heavy); for coding 3, 10 or 30 million per developer; for automation about 1,500 tokens in and 150 out per item.
- The recommendation is the lowest-cost model in the tier we suggest for that use, from the provider you picked if you picked one. Tiers are broad guidance, not quality scores.
- Where a provider offers an introductory price, we use it and show what the same usage will cost once it ends.
- Prices exclude caching and batch discounts, tool fees, long-context premiums, taxes and negotiated rates. Token counts vary by model. Prices and models are rechecked every month.
Use your own key, pick your own model
PC Investigator runs on your own OpenAI, Anthropic or Google key, and you can switch provider or model in Settings at any time.