AI task costs · in pounds
What it costs to read 1,000 scanned invoices a month with AI
DeepSeek V4.1 Flash is the cheapest of the 21 models priced, at 13p for 1,000 invoices a month (0.013p per invoice). The dearest, GPT-6 Astra, costs £19.22; the middle of the range is Gemini 3.8 Flash at £1.44.
How much does it cost to pull supplier, dates and totals out of scanned invoices with a vision AI model? The answer below covers price only, not quality.
Prices of from OpenRouter's public API price list · converted at the ECB reference rate of ($1 = £0.7538) · Urbano DX
Cost by model, cheapest first
The workload
- One scanned A4 page per invoice, sent to the model as an image and counted as 1,500 input tokens. Providers turn images into tokens with their own formulas, so the real count can be several hundred tokens either side.
- 300 tokens of instructions and the list of fields: 1,800 input tokens.
- The answer is supplier, invoice number, dates, net, VAT and total as JSON: 150 output tokens.
- Only models that accept images are listed.
| Model | Per invoice | 1,000 invoices a month | Rate used (USD) |
|---|---|---|---|
| DeepSeek V4.1 FlashDeepSeek | 0.013p | 13p | $0.02016 in, $0.891 out per 1M tokens |
| Gemma 4 31BGoogle | 0.016p | 16p | $0.09 in, $0.34 out per 1M tokens |
| Claude Haiku 5.5Anthropic | 0.019p | 19p | $0.10 in, $0.50 out per 1M tokens |
| GPT-6 LunaOpenAI | 0.019p | 19p | $0.10 in, $0.50 out per 1M tokens |
| Qwen3.8 FlashQwen | 0.026p | 26p | $0.15 in, $0.47 out per 1M tokens |
| Mistral Small 4Mistral | 0.027p | 27p | $0.15 in, $0.60 out per 1M tokens |
| Gemini 2.5 FlashGoogle | 0.069p | 69p | $0.30 in, $2.50 out per 1M tokens |
| Gemini 3.5 Flash LiteGoogle | 0.069p | 69p | $0.30 in, $2.50 out per 1M tokens |
| GPT-4.1 MiniOpenAI | 0.072p | 72p | $0.40 in, $1.60 out per 1M tokens |
| Mistral Large 4Mistral | 0.12p | £1.16 | $0.68 in, $2.09 out per 1M tokens |
| Gemini 3.8 FlashGoogle | 0.14p | £1.44 | $0.75 in, $3.75 out per 1M tokens |
| Claude Haiku 4.5Anthropic | 0.19p | £1.92 | $1.00 in, $5.00 out per 1M tokens |
| Muse Spark 1.3Meta | 0.22p | £2.18 | $1.25 in, $4.25 out per 1M tokens |
| Grok 4.7SpaceXAI | 0.34p | £3.39 | $2.00 in, $6.00 out per 1M tokens |
| Qwen3.8 Max (0902)Qwen | 0.34p | £3.39 | $2.00 in, $6.00 out per 1M tokens |
| Claude Sonnet 5.5Anthropic | 0.38p | £3.84 | $2.00 in, $10.00 out per 1M tokens |
| Gemini 4 ArgonGoogle | 0.38p | £3.84 | $2.00 in, $10.00 out per 1M tokens |
| GPT-6.1 SolOpenAI | 0.38p | £3.84 | $2.00 in, $10.00 out per 1M tokens |
| Claude Opus 5.5Anthropic | 0.77p | £7.69 | $4.00 in, $20.00 out per 1M tokens |
| Claude Fable 5.1Anthropic | 1.9p | £19.22 | $10.00 in, $50.00 out per 1M tokens |
| GPT-6 AstraOpenAI | 1.9p | £19.22 | $10.00 in, $50.00 out per 1M tokens |
How the numbers are worked out
- Each workload is written out on its page: tokens in and out, the share read from a prompt cache, audio minutes or images. Cost = workload × the model's listed rate, so you can check any figure with a calculator.
- Rates are OpenRouter's API prices in our snapshot of , the same ones on our model pricing pages. Buying from the developer directly can cost a little more or less. When rates move, the price change log records it.
- US dollars become pounds at the European Central Bank reference rate of : €1 = $1.1195 = £0.84385, so $1 = £0.7538. VAT, card fees and top-up fees are not included.
- About 0.75 English words make one token. Each model's tokenizer counts the same text a little differently, so treat a figure as a close estimate, not a quote.
- Models that reason before answering can add thinking tokens, billed as output. These figures count none, so a model with thinking switched on can cost more.
- This compares price only. It says nothing about which model does the job well: a cheap model that gets a tenth of the answers wrong can cost more in staff time than it saves.
Other tasks
- Summarise a 50-page PDF
- Draft replies to 1,000 customer emails
- Run a support chatbot: 10,000 chats a month
- Classify 10,000 support tickets
- Transcribe 100 hours of calls
- Write 500 product descriptions
- Translate a 20,000-word website
- Generate 100 marketing images
- Answer questions over 10,000 internal documents
Want this running on your own data, with the cost measured? See the fixed-price AI pilot