About ModelPriceWatch
Verified LLM API pricing, sourced from official pages — not fabricated.
What this is
A data-product website that tracks LLM API pricing across all major providers. Every model has its own page with verified pricing, specifications, a cost calculator, and a direct link to the official pricing source. The homepage is the product: a live, sortable, filterable table of every model's token cost.
Why we built it
The LLM API market moves fast. Prices drop, new models launch, old ones deprecate. Most "pricing comparison" sites either scrape stale data, list fabricated prices, or bury the numbers behind marketing copy. We do the opposite: the data is on the homepage, every price links to its source, and we'd rather list a verified model than a fabricated one.
Affiliate disclosurehow this site pays for itself
Some provider sign-up links on this site are affiliate links. If you sign up through them we may earn a commission at no extra cost to you. This never affects which models we list or how we rank them — pricing data is always shown cheapest-first based on blended cost.
How we verify pricing — and what the verified badge means
Every price in our database is traced to the provider's official pricing page and re-verified on a schedule. When a price is uncertain or we can't confirm it, we exclude the model rather than guess. Each model page links directly to its source pricing URL so you can verify before buying.
The verified badgeprovenance, not endorsement
The badge next to a model means its price is traced to that provider's official pricing page — the source link shown on every model page, so you can check the number yourself in one click — and is re-verified on a schedule, not entered once and forgotten. Two checks run automatically: daily, every price we can map is compared against an independent public reference (the open-source LiteLLM price table) and any divergence is flagged for a human to reconcile against the official page; and weekly, the highest-value models are re-read directly from the provider's own pricing page, capturing a dated receipt. The badge is a provenance signal, not an endorsement — it tells you the number came from, and is reconciled against, the provider's own page, not a third-party guess. If we ever can't tie a price to an official source, we drop the model rather than show an unsourced number.
Official pricing pages we trace toevery price links back to one of these
- OpenAI
- openai.com
- Anthropic
- platform.claude.com
- Google Gemini
- ai.google.dev
- xAI
- docs.x.ai
- DeepSeek
- api-docs.deepseek.com
- Mistral
- mistral.ai
- Amazon Nova
- aws.amazon.com
- Cohere
- cohere.com
- Groq
- groq.com
- Together
- together.ai
How to read the table
Blended cost = (3×input + 1×output) ÷ 4, per 1M tokens. It's a quick comparison metric for mixed input/output workloads.
Methodology note (2026-08-01): the per-model blended figure was re-based from the previous equal-weight average of the two per-token prices to the current 3:1 input-weighted blend ((3×input + 1×output) ÷ 4). Real API workloads read far more tokens than they generate, and the new weighting also matches the LLM Price Index basket. Because output is usually the more expensive side, most blended figures moved down, cheapest-model rankings re-sorted, and a small number of head-to-head comparisons flipped which model reads as cheaper. The underlying input and output prices did not change — only the blend arithmetic did. Figures published before this date used the old 1:1 average.
The colored bar shows each model's blended cost relative to the most expensive model tracked. Price tiers: green (<$1/Mtok blended) = cheap, gold ($1–$15) = mid, red (>$15) = expensive frontier models.
Cached input prices appear under the input price where a provider offers prompt caching — this can cut input costs 50–98% for repeated context. Hosting providers (Groq, Together) re-list open-weight models at their own prices, so you can compare the same model across hosts.
Data sources & methodology
We track 198 models in total from 31 providers. The homepage table shows the 178 you can still buy today; the free JSON API and full dataset return all 198. So if you see 178 on the homepage and 198 in the API, that's intentional — on sale vs. total tracked, not a discrepancy. Prices are per 1M (million) tokens in USD. Open-weight models show the cheapest verified hosted price and are marked with an "Open" badge.
What "current" means here
A model is Current only if the provider's own pricing or model page still listed it on a date we recorded. Every model page shows that date under Availability verified, with a link to the page we checked. If we hold no recent evidence, the page says so in as many words rather than leaving you to assume — an unverified listing is not the same as a live one.
Age is not a lifecycle signal. A model released in 2024 that a provider still sells is Current; a model released last year that has been shut down is not. Today's split across the 198 models we track:
Lifecycle splitas recorded on each model page
- 165 Current — listed on the provider's own page as of the date shown on each model page.
- 1 Preview — listed, but the provider labels it preview or experimental; pricing can change without notice.
- 12 Deprecated — the provider has announced end-of-life. Still purchasable, still in our tables, badged so you don't start new work on one.
- 20 Retired — no longer served on the endpoint we price. Removed from the tables, kept at their own URLs for historical reference, and clearly marked.
We re-check these against provider deprecation pages and correct them in the open. In July 2026 that audit found seven models we had been listing as current which providers had already shut down — the oldest more than a year earlier. Those are now marked Retired with their end dates.
Use the data yourself
All pricing data is available as a free JSON API:
/feeds/models.json— all models with full pricing and specs/feeds/providers.json— all providers
Use it in your tools, dashboards, or cost calculators. No API key required. Full endpoint reference on the API docs page.
Disclaimer
Pricing data is sourced from public documentation and may lag behind official changes. Providers update prices frequently. Always verify current pricing on the provider's official page (linked from every model page) before making purchasing decisions. We are not responsible for pricing decisions made using this data.