Cargando...
Cargando...
We use our own cookies to make the site work, Google cookies to show you advertising, and analytics cookies to find out which pages are read and which are not, in aggregate. If you decline, ads are still shown, but not personalised. You can accept everything or keep only the necessary ones.
Every price on this page was read directly from the provider's own pricing documentation, with the date we checked it and a link to the source. Nothing here is a guess, and no model gets a flattering mark-up.
Prices manually verified on 2026-09-28. Exchange rate: $1 = €0.8763 (Banco Central Europeo, referencia del 23/09/2026, 2026-09-23).
Input and output prices in US dollars per million tokens. "Removed" rows are kept because our articles still reference them, and deleting them would make those explanations unverifiable.
Cached input is billed separately. OpenAI charges a fraction of the normal input rate to read from cache but far more than the normal rate to write to it, so a prompt that gets re-read thousands of times is much cheaper than the headline price suggests.
| Provider | Model | Input / M tokens | Output / M tokens | Context | Max output | Source |
|---|---|---|---|---|---|---|
| OpenAI | GPT-6 AstraOpenAI's first model to reach the "Critical" cybersecurity level under its own Preparedness Framework. Above 272,000 input tokens the price doubles. In Europe, +10% for data residency. | $10.00 | $50.00 | 1,050,000 | 128,000 | OpenAI — Pricing (API)Verified: 2026-09-28 |
| OpenAI | GPT-6 SolReplaced GPT-5.6 Sol on 22 September 2026 at roughly half the price. EU data residency is only available with Standard processing. | $2.00 | $10.00 | 1,050,000 | 128,000 | OpenAI — Pricing (API)Verified: 2026-09-28 |
| OpenAI | GPT-6 LunaIt powers unlimited text chat on ChatGPT's Free tier since August 2026. | $0.10 | $0.50 | 1,050,000 | 128,000 | OpenAI — Pricing (API)Verified: 2026-09-28 |
| Anthropic | Claude Fable 5.1Anthropic's most intelligent generally available model. Cache reads dropped 75% vs Fable 5, cutting typical costs by around 25% and up to 45% for agentic workloads. Cybersecurity or biology queries that trip the safeguards are routed to a less capable model and are not billed at Fable pricing. | $10.00 | $50.00 | 1,000,000 | 128,000 | Anthropic — Claude Platform docsVerified: 2026-09-28 |
| Anthropic | Claude Mythos 5.1Same weights as Fable 5.1 but with relaxed cybersecurity and life-sciences safeguards. Invitation only, via the Cyber Verification Program and the Life Sciences Verification Program (Project Glasswing). Not openly purchasable. | $10.00 | $50.00 | 1,000,000 | 128,000 | Anthropic — Claude Platform docsVerified: 2026-09-28 |
| Anthropic | Claude Opus 5Anthropic recommends starting here for most workloads; Fable 5.1 is reserved for long-horizon agentic work where Opus still falls short. | $5.00 | $25.00 | 1,000,000 | 128,000 | Anthropic — Claude Platform docsVerified: 2026-09-28 |
| Anthropic | Claude Sonnet 5The usual cost/capability balance in the Claude family. | $2.00 | $10.00 | 1,000,000 | 128,000 | Anthropic — Claude Platform docsVerified: 2026-09-28 |
| Anthropic | Claude Haiku 4.5The fastest and cheapest of the current family. 200k context window, half the rest. | $1.00 | $5.00 | 200,000 | 64,000 | Anthropic — Claude Platform docsVerified: 2026-09-28 |
| Gemini 3.8 FlashGoogle's most intelligent Flash model, built for long-horizon software engineering and agents. Output price includes thinking tokens. Flex tier: 0.375 / 1.875. Priority tier: 1.35 / 6.75. | $0.75 | $3.75 | 1,048,576 | 65,536 | Google — Gemini API docsVerified: 2026-09-28 | |
| Gemini 3.5 Flash-LiteGoogle's lowest-cost option for classification, extraction and bulk sorting. | $0.30 | $2.50 | 1,048,576 | 65,536 | Google — Gemini API docsVerified: 2026-09-28 | |
| Meta | Muse Spark 1.3Open model from Meta Superintelligence Labs. The price reflects a partner inference provider, not an official Meta rate: Meta does not publish API pricing for its open models. | $0.20 | $0.80 | 256,000 | 32,000 | Meta AI — ModelosVerified: 2026-09-28 |
| Meta | Llama 6Open model from Meta. Served through third-party inference providers, so the final price depends on the provider you pick. | $0.20 | $0.80 | 256,000 | 32,000 | Meta AI — LlamaVerified: 2026-09-28 |
| DeepSeek | DeepSeek Chat (V3)The lowest price per token in the catalogue, at the cost of a much older knowledge cutoff and a shorter context window. Good for bulk classification and simple high-volume work. | $0.28 | $0.42 | 128,000 | 16,000 | DeepSeek — PricingVerified: 2026-09-28 |
What you actually pay per month if you are a person rather than a developer. API billing and subscriptions are not comparable: one is metered, the other is flat.
OpenAI
$0 per month
OpenAI
$8 per month
OpenAI
$20 per month
OpenAI
$100 per month
OpenAI
$200 per month
OpenAI
$25 per month
per year: $20 per month
OpenAI
$125 per month
per year: $100 per month
Anthropic
$20 per month
Anthropic
$100 per month
Anthropic
$200 per month
$19.99 per month
Midjourney
$10 per month
per year: $8 per month
Midjourney
$30 per month
per year: $24 per month
Midjourney
$60 per month
per year: $48 per month
Midjourney
$120 per month
per year: $96 per month
Per image, or per second of video. Watch out for the billing unit: some providers bill the whole clip even if you only use four seconds of it.
| Model | Price | What it includes | Source |
|---|---|---|---|
| Imagen — calidad estándarOpenAI · GPT Image (vía API) | $0.04 |
| OpenAI — Pricing (API, imágenes)Verified: 2026-09-28 |
| Imagen — calidad altaOpenAI · GPT Image (vía API) | $0.17 |
| OpenAI — Pricing (API, imágenes)Verified: 2026-09-28 |
| Vídeo — SoraOpenAI · Sora 2 | $0.5/s |
| OpenAI — SoraVerified: 2026-09-28 |
| Vídeo — VeoGoogle · Veo 3 | $0.35/s |
| Google — Veo API pricingVerified: 2026-09-28 |
| Imagen — Imagen / Nano BananaGoogle · Gemini 3 Pro Image | $0.04 |
| Google — Gemini API pricingVerified: 2026-09-28 |
Prices per million tokens are only half the answer. If you know how many tokens your own task uses, this works out the actual cost per run and per month.
Quick example
1,050,000-token context · max 128,000 output · knowledge cutoff 2026-04-30
1,200
0
400
Cost of one run
€0.028
= $0.03 · 1,600 tokens in total
Monthly cost (1,000 runs)
€28.0431
$32.00
With $100.00 per month
3,125
runs per month
Price source
OpenAI — Pricing (API)Verified: 2026-09-28
Exchange rate applied: 1 $ = 0.8763 € (Banco Central Europeo, referencia del 23/09/2026, 2026-09-23)