pricing
Pay as you go for EU-hosted inference. No subscription fees, no hidden platform charge, just prepaid credits.
€0
no monthly fee
create an account for free
pay as you go
usage-based billing
top up credits only when you need them
€5
minimum top-up
iDEAL, cards, Bancontact, and more
web search
The optional hosted web search feature (off by default) bills a flat per-search fee from your prepaid credits. The Research button runs one deep search (multi-step, 5 to 30 seconds) at a separate, higher fee. Both prices are shown in the chat composer's capabilities menu.
Queries are processed by Linkup in the EU under zero data retention. A self-hosted search MCP server is free to connect.
per-model pricing
these are the prices you pay on ozeye, ozeye's margin included. most models run at several EU providers, and you pay the price of the provider that serves your request. where providers differ, the table shows the range from the cheapest to the most expensive. each model page lists every provider's price, the usage dashboard shows which provider served each request, and you can block providers in your settings. a dash for cached input means it costs the same as input.
| model | input / 1M | cached / 1M | output / 1M | providers |
|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | €0.1245to €0.4480 | €0.0287to €0.0896 | €0.2682to €0.8960 | 3 |
| DeepSeek V4 Pro 0813 | €1.916 | €0.4790 | €3.832 | 1 |
| DeepSeek V4.1 Flash | €0.4790 | €0.1245 | €1.437 | 1 |
| GLM 5.2 | €1.150to €2.016 | €0.2491to €2.016 | €4.023to €6.160 | 4 |
| GLM 5.3 | €1.341to €1.676 | €0.1341to €0.4191 | €4.215to €4.311 | 3 |
| GLM 5.3 Flash | €0.1437to €0.1916 | €0.0479to €0.0671 | €0.4790 | 2 |
| Kimi K3 | €2.874 | €0.7185 | €14.370 | 1 |
| MiniMax M3 | €0.3832 | €0.0958 | €1.916 | 1 |
| Kimi K2.6 | €0.6993 | €0.2395 | €3.353 | 1 |
| Kimi K2.7 Code | €0.7185to €1.198 | €0.1916to €0.2994 | €3.353to €4.311 | 2 |
| Magistral Medium | €1.916 | - | €4.790 | 1 |
| Qwen 3.8 2.4T A95B | €2.395 | €0.5987 | €5.748 | 1 |
| Qwen 3.8 27B | €0.3832to €0.6720 | €0.0958to €0.4480 | €2.299to €3.696 | 3 |
| Qwen 3.8 Flash Next | €0.1916 | €0.0479 | €0.4790 | 1 |
| Qwen 3.6 35B | €0.2800 | - | €1.680 | 1 |
| MiniMax M2.5 | €0.2874 | €0.0718 | €1.150 | 1 |
| Gemma 4 26B | €0.2800 | - | €0.5600 | 1 |
| GPT-OSS 120B | €0.0896to €0.1680 | - | €0.4480to €0.6720 | 2 |
| GPT-OSS 20B | €0.0448 | - | €0.1680 | 1 |
| LLaMA 3.3 70B | €0.7504to €1.008 | - | €0.7504to €1.008 | 2 |
| Magistral Small | €0.4790 | - | €1.437 | 1 |
| Ministral 3 14B | €0.1916 | - | €0.1916 | 1 |
| Ministral 3 3B | €0.0958 | - | €0.0958 | 1 |
| Ministral 3 8B | €0.1437 | - | €0.1437 | 1 |
| Mistral Large | €0.4790 | - | €1.437 | 1 |
| Mistral Medium | €1.437 | - | €7.185 | 1 |
| Mistral Medium 3.5 | €1.680 | - | €8.400 | 1 |
| Mistral Small | €0.1437 | - | €0.5748 | 1 |
| Mistral Small 3.2 Vision | €0.0958to €0.1680 | - | €0.2969to €0.3920 | 2 |
| Mistral Small 4 | €0.1008 | - | €0.3024 | 1 |
| Pixtral 12B | €0.2240 | - | €0.2240 | 1 |
| Qwen 3 235B | €0.8400 | - | €2.520 | 1 |
| Qwen 3 Coder 30B | €0.0671to €0.2240 | - | €0.2491to €0.8960 | 2 |
| Qwen 3.5 397B | €0.6720 | - | €4.032 | 2 |
| Qwen 3.5 9B | €0.1120to €0.1437 | €0.0359to €0.1120 | €0.1680to €0.1916 | 2 |
| Qwen 3.6 27B | €0.4480 | - | €3.024 | 1 |
| Voxtral Small | €0.0958 | - | €0.3832 | 1 |
| Mistral 7B v0.3 | €0.1054 | - | €0.1054 | 1 |
| Mistral OCR | €0.7168 | - | €0.7168 | 1 |
| Mistral OCR v4 | €0.7168 | - | €0.7168 | 1 |
| Codestral | €0.2874 | - | €0.8622 | 1 |
| Qwen 2.5 VL 72B | €1.019 | - | €1.019 | 1 |
| Codestral Embed | €0.1437 | - | - | 1 |
| BGE M3 | €0.0112 | - | - | 1 |
| BGE Multilingual Gemma2 | €0.0112to €0.1120 | - | - | 2 |
| Qwen 3 Embedding 8B | €0.0192to €0.1120 | - | - | 3 |
| Mistral Embed | €0.0958 | - | - | 1 |
| Mistral Moderation | free | 1 | ||
| Qwen3 Guard 0.6B | free | 1 | ||
| Qwen3 Guard 8B | free | 1 | ||
audio, speech and images
| model | price | unit |
|---|---|---|
| Voxtral Mini Realtime | €0.0059 | per minute of audio |
| Voxtral Mini Transcribe 2 | €0.0029 | per minute of audio |
| Faster Whisper Large V3 | €0.0038 | per minute of audio |
| Whisper Large V3 | €0.0027to €0.0034 | per minute of audio |
| Whisper Large V3 Turbo | €0.0009 | per minute of audio |
| Chatterbox Turbo | €4.790 | per 1M characters |
| Voxtral Mini TTS | €15.328 | per 1M characters |
| Stable Diffusion XL | €0.0100 | per image |
frequently asked questions
how does billing work?
You add credits to your account and each request deducts usage from that balance based on the selected model and tokens processed.
which price do I pay when a model has several providers?
You pay the price of the provider that serves your request, plus ozeye's margin. Which provider serves depends on routing and availability, and a request can fail over to another provider. The table above shows the range from the cheapest to the most expensive provider, each model page lists every provider's price, and the usage dashboard shows which provider served each request. You can block providers you do not want in your settings.
what is the minimum top-up?
The minimum credit purchase is €5. Regular new accounts start with €0 credits; beta credits are only granted to eligible beta signups.
can I see exactly what I'm paying?
Yes. The usage dashboard shows per-request costs and you can export detailed usage data as CSV or JSON.
do I need a card to create an account?
No. Creating an account is free; paid usage requires prepaid credits.