“Which is cheaper, GPT, Claude, or Gemini?” has no single answer, because each lab ships a ladder of models, and the three ladders don’t line up rung for rung. The honest comparison is tier by tier. Here’s how the three families stack up as of our 2026-07-14 snapshot, using each model’s separate input and output price plus a blended figure (a 3:1 input-to-output mix, (3×in + out)/4) for quick ranking.
Flagship tier
The top-end, high-capability models:
| Model | Input / 1M | Output / 1M | Blended | Context |
|---|---|---|---|---|
| Claude Opus 4.8 | $5.00 | $25.00 | $10.00 | 1M |
| GPT-5.6 Terra | $2.50 | $15.00 | $5.63 | 1.05M |
| Gemini 3.1 Pro Preview | $2.00 | $12.00 | $4.50 | 1.05M |
At the top, Gemini undercuts GPT undercuts Claude, and the spread is wide: Claude Opus 4.8 costs more than twice Gemini 3.1 Pro on a blended basis. Anthropic positions Opus as a premium reasoning flagship and prices it like one. If your flagship workload is output-heavy, the gap widens further — Opus’s $25 output is more than double Gemini’s $12.
That doesn’t make Opus the wrong choice; it makes it a choice you should only pay for when the task needs it. Many teams route the hard 10% of requests to a flagship and everything else down a tier.
Mid tier — the workhorses
Where most production traffic actually runs:
| Model | Input / 1M | Output / 1M | Blended | Context |
|---|---|---|---|---|
| Claude Sonnet 5 | $2.00 | $10.00 | $4.00 | 1M |
| Gemini 3.5 Flash | $1.50 | $9.00 | $3.38 | 1.05M |
| GPT-5.6 Luna | $1.00 | $6.00 | $2.25 | 1.05M |
Here the order flips: GPT-5.6 Luna is the cheapest of the three, roughly 44% below Claude Sonnet 5 on blended cost, with Gemini 3.5 Flash in between. This is the tier to study hardest, because it’s where you’ll spend the most money over a year and where the models are closest in capability. A blended range of $2.25–$4.00 means model choice alone swings a mid-tier bill by ~75%.
Budget tier
For classification, routing, extraction, and high-volume simple tasks:
| Model | Input / 1M | Output / 1M | Blended | Context |
|---|---|---|---|---|
| Claude Haiku 4.5 | $1.00 | $5.00 | $2.00 | 200K |
| Gemini 3.1 Flash Lite | $0.25 | $1.50 | $0.56 | 1.05M |
OpenAI’s smallest current tier prices close to Luna in our data, but the standout here is Gemini 3.1 Flash Lite: at $0.56 blended it’s roughly a quarter of Claude Haiku’s cost and carries a million-token context. For pure volume work where you don’t need frontier reasoning, Google’s budget rung is aggressively cheap. Note the context difference too — Haiku’s 200K window is generous for chat but a fraction of the Gemini and GPT million-token windows.
Two things the headline price hides
Context window. Claude’s flagship and mid models cap at 1M tokens; Google and OpenAI push to ~1.05M across most tiers, and Anthropic’s budget Haiku drops to 200K. If your workload depends on stuffing large documents into a single call, the cheaper model with the bigger window can win twice.
Cache rates. All three families discount cached input heavily — around 90% off on the models here (Claude Sonnet 5 caches at $0.20 vs $2.00 fresh; GPT-5.6 Luna at $0.10 vs $1.00; Gemini 3.1 Pro at $0.20 vs $2.00). If your prompts reuse a big fixed prefix, the cached input price matters more than the sticker input price. A model that looks pricier fresh can be cheaper in practice once caching kicks in.
So which is cheapest?
- Flagship: Gemini < GPT < Claude (Claude Opus is the premium option).
- Mid tier: GPT < Gemini < Claude.
- Budget: Gemini Flash Lite is in a class of its own on price.
Google is the most consistently aggressive on price across tiers; OpenAI wins the mid tier; Anthropic prices for the top end and asks you to pay for it. But “cheapest blended” is a starting filter, not a verdict — your real input:output ratio, your context size, your cache hit rate, and (above all) which model actually solves your task will move the answer.
Every price here is from a single day, 2026-07-14, and these numbers move — in the two weeks before this snapshot, one major model cut its rates by a third. Each model on computetrail carries its own daily price history, so you can check whether these standings still hold before you build on one.
Figures from computetrail’s 2026-07-14 price snapshot; blended figures assume a 3:1 input-to-output mix and are for ranking, not budgeting. Independent and not affiliated with any provider. See our methodology.