Volume
flat rate at scale
The problemThousands of pieces and variations a month make premium per-token pricing unviable.
HelmcodeA flat rate per key: produce as much as the campaign needs at a fixed cost.
industries · Media & agencies
Produce at volume. Pay a flat rate.
Thousands of pieces a month make premium per-token pricing unworkable. Open models on a flat rate, with the best open writing in Spanish, make volume affordable.
the economics
The economics, by design.
Each cost lever mapped to a platform capability. Sovereignty comes by default.
Spanish
Qwen 3.6, best open ES writing
The problemAgencies working in Spanish need native-quality copy, not translated English.
HelmcodeQwen 3.6 is the strongest open model for Spanish writing, tone and brand voice.
Sovereignty
EU · client data
The problemClient briefs and unpublished campaigns are confidential.
HelmcodeEU-only inference, zero logs: client material never trains anyone’s model.
what has to carry a label
The labelling rules for generated work.
If you produce content at volume, the part of the AI Act that lands on you is Article 50, and it is more specific than the debate around it suggests. Two of its exceptions are worth knowing before you write a policy for the studio.
01
Machine-readable, not just visible
Article 50(2) asks that synthetic audio, image, video or text be marked in a machine-readable format and be detectable as artificially generated or manipulated. A watermark a human can see is not the same as metadata a platform can read, and the text asks for the second.
02
Deepfakes, with an art exception
Paragraph 4 makes the deployer disclose that content is artificially generated or manipulated. For evidently artistic, creative, satirical or fictional work the duty softens: disclosure happens in an appropriate manner that does not hamper the display or enjoyment of the work.
03
Editorial responsibility replaces the label
For AI-generated text published to inform the public on matters of public interest, disclosure is required, except where a natural or legal person holds editorial responsibility for the publication. That exception is the one an agency should read twice, because it is about who signs, not about how the text was made.
EU AI Act · Regulation (EU) 2024/1689 Regulation (EU) 2024/1689 of 13 June 2024, Article 50, paragraphs 2 and 4. These transparency duties sit outside the high-risk regime and apply on their own terms. How they interact with your clients’ own advertising and platform obligations is a separate question. read the report →
use cases
Your most common use cases.
The cases with the most traction in the sector, each with its own page in detail.
Content generation Copy, variations and brand content at volume, in native Spanish. Translation Bulk translation across markets at marginal cost. Semantic search & recommendation Semantic search over assets, briefs and past campaigns. Voice & transcription Voiceover and audio content with an open pipeline. Summarization Briefs, coverage reports and research decks condensed for a client who has ten minutes. Classification & triage Tagging a media library or a campaign backlog so the asset you need is findable at all.
Recommended open models.
A starting point per task type. The full guide maps 80 cases to the open model for each one.
- Qwen 3.6 in Helmcode
- The best open writing in Spanish: copy, brand voice, adaptation.
- Qwen3.6-27B
- Dozens of A/B variations per campaign at zero marginal cost.
- DeepSeek V4 Flash in Helmcode
- Bulk translation and high-volume generation at the lowest cost.
We are distilling and quantizing these open models into small, tightly specialised versions, trained for one task rather than for all of them. A model like that runs on less hardware, answers faster and fits where the big one does not, your own datacenter included. If you have a process with volume and stable criteria, that is the conversation we want to have with you.
// faq
Questions, answered.
What the sector's technical, compliance and business teams ask.
Why does per-token pricing hurt at content volume?
Producing thousands of pieces and variations a month means the bill scales linearly with output. A flat rate per key removes that ceiling: you produce as much as the campaign needs at a fixed cost.
How good is the Spanish?
Qwen 3.6 is the strongest open model for Spanish writing, which is the deciding argument for agencies working in Spanish, and it runs privately so client briefs never leave.
Is client material kept confidential?
Yes. EU-only inference with zero logs: briefs, unpublished campaigns and assets are never stored and never train a model.
Can we generate many variations at once?
Yes. Qwen3.6-27B produces dozens of A/B variations per campaign at near-zero marginal cost, on a flat rate.
// get started
START BURNING TOKENS
Skip the AI infra work. Deploy your first private inference endpoint today.
Flat rate. EU data. OpenAI API compatible.