docs/models.md

Models

The co-op serves models from multiple providers. This is the single list — both the chat and the API draw from it, so it's the one place to look (and the one place to update).

Providers

Models are named provider/model-name, so you can choose not just which model but where it runs. Providers differ along two independent axes: privacy (how protected your prompts and responses are) and green energy (how the compute is powered). These are gradations, not binary switches — and no provider today is best on both.

Provider	Value	Privacy	Green energy	What that means
Tinfoil	private	Strongest — architectural	Not a green claim	Runs in hardware enclaves (TEEs) with end-to-end encryption (EHBP), verifiable via remote attestation. Neither we nor Tinfoil can read your prompts or responses at inference time. Energy mix is not disclosed.
GreenPT	green	Policy-based	Strongest — 100% renewable	Standard (non-enclave) hosting in the EU; privacy rests on GreenPT's commitments and EU data rules, not on cryptography. Compute runs on 100% renewable energy.
Regolo	green	Policy-based — zero data retention	100% renewable	EU-hosted (Italy, Seeweb) with a stated zero-data-retention policy: prompts are not stored or reused. Compute runs on 100% renewable energy.

Reading the gradations:

• Privacy: architectural (Tinfoil) beats policy-based (GreenPT, Regolo). An enclave protects you even if the provider is compromised; a policy protects you only as far as the provider keeps its word. Both are better than a provider with neither — and Regolo adds a written zero-data-retention commitment on top of its policy. How the co-op measures its own energy use and carbon emissions is on the energy and carbon page — including what those figures cover and what they miss.

• Green energy: 100% renewable (GreenPT, Regolo) beats undisclosed (Tinfoil). GreenPT and Regolo both run on 100% renewable energy; Tinfoil's mix is not disclosed.

Trade-offs are real: the most private option (Tinfoil) isn't the greenest, and the greenest option (GreenPT) isn't the most private. The point of the co-op is that you choose which axis matters more for a given task.

How to pick, quick guide:

• Default → greenpt/deepseek-v4.1-flash: DeepSeek's efficient multimodal reasoning model on renewable energy, and the model the chat starts on. For the TEE-protected (private) version of the same model, use tinfoil/deepseek-v4-1-flash.
• Everyday chat and drafting → a fast model: greenpt/glm-5.3-flash. These handle the bulk of what people use chat for, at a fraction of flagship cost.
• Hard reasoning, coding, agentic work → a flagship: greenpt/glm-5.3 (renewable) or tinfoil/glm-5.3 (TEE-protected) — the GLM flagship on the provider of your choice — or tinfoil/deepseek-v4-1-flash for long documents and tool-driven tasks.
• Images in, text out (vision) → tinfoil/deepseek-v4-1-flash or the Apertus models — see the ✅ in the table below.
• Cheap and cheerful → tinfoil/gpt-oss-120b or greenpt/glm-5.3-flash.
• Model origins matter to you → the Apertus models (publicly developed; see below).
• Semantic search / RAG (API) → an embedding model: greenpt/qwen3-embedding-8b or greenpt/green-embeddings (renewable energy), or tinfoil/nomic-embed-text (TEE-protected — the model the co-op's own docs search runs on). API-only; see Embedding models below.

Current models

Pricing is per 1 million tokens, input / output / cached input (what the co-op pays the provider — your own cost is your monthly allowance, not per-token billing). The cached-input rate applies when a model reuses a shared prompt prefix (multi-turn conversations, agent loops); most providers charge a steep discount for it. Models with — don't offer a distinct cache-read price. Vision means the model accepts images as input (screenshots, photos, documents to reason about). Image generation is available via the chat's image feature (the 🎨 button) using regolo/qwen-image.

The emoji in the model name marks its value tier (the same icons used in the chat picker):

• 🔒 Private — Tinfoil: architecturally private (TEE enclaves)
• 🌱 Green — GreenPT and Regolo: renewable-powered EU hosts (Regolo adds a zero-data-retention commitment). The 🌱 emoji marks renewable-powered providers, not a specific company.

Model	Provider	Privacy	Green	Vision	Context	Pricing (in / out / cached)	Notes
🌱 greenpt/deepseek-v4.1-flash	GreenPT	policy	✅ 100% renewable	✅	1M	$0.22 / $1.10 / $0.011	Default. DeepSeek's multimodal reasoning model, renewable energy; TEE twin below. Base V4.1 Flash release — not GreenPT's 17 Oct 2026 "1017" snapshot (as of 2026-11-09).
🔒 tinfoil/deepseek-v4-1-flash	Tinfoil	architectural	undisclosed	✅	1M	$0.65 / $1.45 / $0.13	TEE-protected version of the default; image input, tool calling, long docs/agents.
🔒 tinfoil/glm-5.3	Tinfoil	architectural	undisclosed	—	1M	$1.80 / $5.75 / $0.45	GLM flagship, TEE-protected; strongest private-tier model for agentic coding and long-horizon tasks.
🔒 tinfoil/gpt-oss-120b	Tinfoil	architectural	undisclosed	—	131K	$0.15 / $0.60 / —	Lightweight open-weight reasoning; cheapest private tier.
🌱 greenpt/glm-5.3-flash	GreenPT	policy	✅ 100% renewable	—	1M	$0.11 / $0.44 / $0.022	Fast GLM model, renewable energy.
🌱 greenpt/glm-5.3	GreenPT	policy	✅ 100% renewable	—	1M	$1.10 / $4.40 / $0.275	GLM flagship — strongest agentic model in the green tier.
🌱 regolo/apertus-70b	Regolo	policy + ZDR	✅ 100% renewable	✅	262K	$1.16 / $3.94 / —	Fully open sovereign model (Apache-2.0), large/capable, image understanding.
🌱 regolo/qwen3.5-122b	Regolo	policy + ZDR	✅ 100% renewable	✅	262K	$0.70 / $3.94 / —	Large vision+reasoning MoE (Qwen 3.5, 122B-A10B); 262K context.

Vision is verified working through the co-op gateway (base64 image input via /v1/chat/completions, standard OpenAI format). In the chat UI, just attach an image to your message while using a ✅ model.

Context figures are measured the same way, through the gateway. Where a provider's public catalog advertises a smaller limit than the model actually accepts, the measured figure is the one shown here.

Reading the Privacy column: architectural means the protection is built into the infrastructure (TEE enclaves) — even the provider can't read your prompts. Policy means the provider commits, by contract and policy, not to store or reuse your data. ZDR (zero data retention) is a stronger policy: the provider keeps no copy of prompts or outputs at all after the response is delivered.

Same model, two providers: glm-5.3 (Z.ai's GLM flagship) is offered both on GreenPT (renewable, $1.10/$4.40) and on Tinfoil (TEE-protected, $1.80/$5.75) — pick the axis that matters to you. The fast glm-5.3-flash variant is GreenPT-only (Tinfoil deprecated their flash variant in October 2026).

Audio models

These are not chat models — they power the chat's voice mode (dictation and spoken replies) and are available through the API on /v1/audio/* routes for members building speech into their own tools. They're hidden from the chat model picker to avoid confusion.

Model	Provider	Use	Pricing
🔒 tinfoil/whisper-large-v3-turbo	Tinfoil	Speech-to-text	$0.01 per request (flat; zero per-token)
🔒 tinfoil/voxtral-small-24b	Tinfoil	Speech-to-text (token-priced)	$0.20 / 1M input + $0.60 / 1M output tokens
🔒 tinfoil/voxtral-tts	Tinfoil	Text-to-speech	$0.002 per request (zero per-token)

All three run inside Tinfoil's enclaves, so voice data gets the same architectural privacy as the private chat models. TTS voices available: neutral_female, neutral_male, casual_female/male, cheerful_female, plus French, German, Spanish, Italian, Portuguese, Dutch, and Hindi variants.

Tinfoil bills these at cost, but its audio API returns no token usage to the gateway — so dashboard rows for these models read $0.01 · 0 tokens: the charge is real, the token columns stay empty.

whisper-large-v3-turbo and voxtral-tts charge flat per request. voxtral-small-24b is token-priced and the better pick for small requests: sent through the chat API with input_audio it meters exactly per token, and on /v1/audio/transcriptions (the route dictation uses) the gateway bills Tinfoil's measured chunk-equivalent — their audio encoder tokenizes in ~30 s chunks, so a short request costs ≈$0.0001, about 1/125th of the whisper floor. For audio longer than ~30 s, prefer the chat route, which meters exactly.

# Speech-to-text
curl https://gateway.inference.coop/v1/audio/transcriptions \
  -H "Authorization: Bearer ***" \
  -F file=@your-audio.mp3 \
  -F model=tinfoil/whisper-large-v3-turbo

# Speech-to-text (token-priced model, chat route — meters exactly per token)
curl https://gateway.inference.coop/v1/chat/completions \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{"model": "tinfoil/voxtral-small-24b", "messages": [{"role": "user", "content": [
        {"type": "text", "text": "Transcribe this audio."},
        {"type": "input_audio", "input_audio": {"data": "<base64 audio>", "format": "wav"}}
      ]}]}'

# Text-to-speech (returns WAV audio)
curl https://gateway.inference.coop/v1/audio/speech \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{"model": "tinfoil/voxtral-tts", "input": "Hello from the co-op", "voice": "neutral_female"}' \
  --output speech.wav

# Vision (image input to a multimodal chat model)
curl https://gateway.inference.coop/v1/chat/completions \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{"model": "tinfoil/deepseek-v4-1-flash", "messages": [{"role": "user", "content": [
        {"type": "text", "text": "What is in this image?"},
        {"type": "image_url", "image_url": {"url": "data:image/png;base64,..."}}
      ]}]}'

Embedding models

Embeddings turn text into a numeric vector — the input to semantic search, clustering and RAG. They're called on /v1/embeddings with your co-op API key and are not chat models: they never appear in the chat picker, the API is the interface (the same shape as the audio models above). You choose between them the same way as for chat: a TEE-protected option (Tinfoil) or renewable-energy options (GreenPT).

Model	Provider	Privacy	Green	Dimensions	Max input	Pricing
🌱 greenpt/green-embeddings	GreenPT	policy	✅ 100% renewable	2560 (measured)	32K tokens (per GreenPT's catalog)	$0.20 / 1M input tokens (no output charge)
🌱 greenpt/qwen3-embedding-8b	GreenPT	policy	✅ 100% renewable	4096 (measured)	32K tokens (per GreenPT's catalog)	$0.25 / 1M input tokens (no output charge)
🔒 tinfoil/nomic-embed-text	Tinfoil	architectural (TEE)	— (mix not disclosed)	768 (measured)	8K tokens (per Tinfoil's catalog)	$0.016 / 1M input tokens (no output charge)

curl https://gateway.inference.coop/v1/embeddings \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{"model": "tinfoil/nomic-embed-text", "input": "the text to embed"}'

The response is standard OpenAI shape: data[0].embedding holds the vector, and usage.prompt_tokens is what meters against your allowance at the rate above. Embeddings bill input tokens only — there is no output charge.

Vectors from different models are not interchangeable — pick one per search index and stay with it (re-index if you switch). All three are callable with a member key; for reference, the co-op's own documentation search runs on tinfoil/nomic-embed-text inside the TEE.

Image generation

🌱 regolo/qwen-image (Regolo, EU-hosted, zero data retention) generates images from a text prompt. It runs the December 2025 "2512" checkpoint of Qwen-Image — not the newer Qwen Image 2.1 (September 2026), which Regolo does not list (as of 2026-11-09). In the chat, press the Integrations button below the message box, choose Image, and describe what you want — the generated image appears in the conversation. Via the API:

curl https://gateway.inference.coop/v1/images/generations \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{"model": "regolo/qwen-image", "prompt": "a watercolor painting of a mountain lake", "n": 1, "size": "512x512"}'

The response returns base64-encoded JPEG(s). Billed per pixel generated.

About the Apertus models

Apertus is the Swiss AI Initiative's open foundation model, built by a public collaboration between EPFL, ETH Zurich, and the Swiss National Supercomputing Centre (CSCS). It's "fully open" in the strict sense — not just downloadable weights:

• Open weights & code — released under the Apache-2.0 license, so anyone can use, modify, and redistribute it, commercially or not.
• Open training data — trained only on fully open data (unlike most proprietary models, which keep their training set secret). The data, code, methods, and alignment principles are all published and reproducible.
• Open values — governed by a public Apertus Charter that documents the model's values and principles rather than leaving them implicit.
• Sovereign & public-good — designed to European data-protection norms (GDPR, EU AI Act, Swiss law) as an example of "AI as a public good," not a proprietary product. Trained across 1,500+ languages (40% non-English).

The v1.5 generation adds image understanding, a longer 262K-token context window, improved tool use, and instruction-following. It's the publicly developed counterpart to the private (Tinfoil) and renewable (GreenPT) offerings — sovereignty as a third axis of member choice.

For more information on open-source AI and the Apertus models, see the European Open Source AI Index.

Source: apertus-ai.org and the Swiss AI Hugging Face model cards.

Some models also have a discounted cached-input rate (when a prompt is reused and doesn't need to be re-read) — see the code/litellm config for the exact numbers. Pricing here mirrors the model_info in the LiteLLM config (code/litellm/config.yaml), which is the authoritative source — the two are kept in sync.

Decisions

Most models here generate text. Metis, from GreenPT, makes decisions instead: you give it a state (the situation, in plain text) plus up to 32 typed questions, and it returns probabilities, choices, and scores — never prose. Use it when you want a calibrated judgement rather than a written answer: "should we…?", "which option…?", "how confident is that?".

It's on the gateway at /v1/systemone, called with your normal co-op API key (the same key you use for chat):

curl https://gateway.inference.coop/v1/systemone \
  -H "Authorization: Bearer ***" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "metis",
    "state": "Your situation, described in plain text.",
    "questions": {
      "go":         {"type": "noul", "instructions": "Should we proceed?",
                     "criteria": {"true": "proceed", "false": "do not proceed"}},
      "which":      {"type": "choice", "instructions": "Which factor matters most?",
                     "criteria": {"cost": "price", "speed": "latency"}},
      "confidence": {"type": "score", "instructions": "How confident are you?",
                     "criteria": ["very low", "low", "medium", "high", "very high"]}
    }
  }'

Metis is hosted by GreenPT — the same policy-based privacy, EU hosting, and 100% renewable energy as the 🌱 chat models. Decisions count toward your monthly allowance, like every other call: each request is metered at $0.12 per 1M input tokens, computed from the exact token count GreenPT reports back, and appears in your usage record alongside your chat calls. (Metis generates no prose, so there is no output charge.)

Limitations on privacy

Only Tinfoil offers architectural privacy. The other providers are chosen for their value (renewable energy, public models) but do not run in enclaves — prompts and responses pass through them in the ordinary way. Your chat history is also stored on our server so you can revisit it, and that stored history is not encrypted in a way that prevents us from technically reading it — we commit not to. Requests made through the API with your own key are a different matter: those are not stored on our side at all, beyond the usage record needed for metering (API access). The full distinction — what's architecturally private versus what's a policy commitment — is in the Privacy Policy.