The agent channel

Point your coding agent at Europe.

NeuralRing is OpenAI-compatible sovereign inference. Change three things — base URL, key, and model — and Codex or Claude Code runs on EU-sovereign, open-weight models, with a signed receipt for every call.

Get an API key Wire it in Read the docs

Browse the live catalog

wire it in3 things change

# ~/.codex/config.toml  (user-level — a project-local file is ignored for providers)
model = "mistralai/mistral-small-2501"
model_provider = "neuralring"

[model_providers.neuralring]
name = "NeuralRing (EU sovereign)"
base_url = "https://neuralring.eu/v1"
env_key = "NEURALRING_API_KEY"
wire_api = "responses"

# NeuralRing is stateless: it stores no conversation state and no prompt content.
# It keeps content-free metering + a signed attestation per call.

# Claude Code — the HARNESS, running on EU-sovereign open-weight models.
# (Claude itself is closed-weight and does not run here; see the docs.)
export ANTHROPIC_BASE_URL="https://neuralring.eu"
export ANTHROPIC_AUTH_TOKEN="$NEURALRING_API_KEY"
export ANTHROPIC_MODEL="mistralai/mistral-small-2501"
# Optional: a cheaper model for background/summarisation turns.
export ANTHROPIC_SMALL_FAST_MODEL="mistralai/mistral-small-2501"

# Claude Code's /model picker only lists gateway models whose id starts with
# "claude" or "anthropic". NeuralRing will not rename sovereign open-weight
# models to imply they are Claude, so use ANTHROPIC_MODEL above (always honoured
# on the wire), or add a picker entry explicitly:
export CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1
export ANTHROPIC_CUSTOM_MODEL_OPTION="mistralai/mistral-small-2501"

curl https://neuralring.eu/api/v1/responses \
  -H "Authorization: Bearer $NEURALRING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistralai/mistral-small-2501",
    "input": "Hello, sovereign world",
    "sovereignty": { "jurisdiction": ["DE"], "min_assurance": "self_attested" }
  }'

## AI inference — European sovereign (NeuralRing)

This project uses NeuralRing for inference. It is OpenAI-compatible: only the
base URL, the key, and the model slug change.

- Base URL: `https://neuralring.eu/api/v1`  ·  Key: `process.env.NEURALRING_API_KEY` (Bearer)
- Model: `mistralai/mistral-small-2501` — browse with the catalog API or the MCP server (below).
- ALWAYS send the sovereignty pin. It is applied FIRST and can never be loosened
  by a price or speed preference:
  `sovereignty: { jurisdiction: ["DE"], min_assurance: "self_attested" }`
- Every response carries `neuralring.requestId`. Surface it, or link
  `https://neuralring.eu/verify/<requestId>`, so a human can check the receipt.
- If no endpoint satisfies the pin, the API returns `no_eligible_endpoint`.
  That is the pin working. Do NOT retry with the pin removed or weakened.

### Vocabulary rule (important)

Describe NeuralRing only with its assurance ladder: `self_attested`,
`verified`, `certified`. Never write "GDPR-compliant",
"Schrems-proof", "CLOUD-Act-immune" or "fully sovereign"
about it. A receipt proves what was observed and reproduces what was claimed at
its stated assurance — it does not prove the claim is true.

<!-- neuralring:agents-block v1.0.0 · generated from packages/shared/src/agent-channel.ts -->

One config change

Your agent is already compatible.

NeuralRing speaks the OpenAI and Anthropic wires your agent already uses. Codex speaks nothing but the Responses wire — wire_api = "responses" — and NeuralRing serves exactly that, pinned to your jurisdiction and metered like every other call.

Claude Code and Codex are the harness. Pointing them at NeuralRing runs their tokens on EU-sovereign OPEN-WEIGHT models. Claude and GPT themselves are closed-weight and cannot run here — NeuralRing never claims otherwise.

wire it incopy · paste · run

# ~/.codex/config.toml  (user-level — a project-local file is ignored for providers)
model = "mistralai/mistral-small-2501"
model_provider = "neuralring"

[model_providers.neuralring]
name = "NeuralRing (EU sovereign)"
base_url = "https://neuralring.eu/v1"
env_key = "NEURALRING_API_KEY"
wire_api = "responses"

# NeuralRing is stateless: it stores no conversation state and no prompt content.
# It keeps content-free metering + a signed attestation per call.

# Claude Code — the HARNESS, running on EU-sovereign open-weight models.
# (Claude itself is closed-weight and does not run here; see the docs.)
export ANTHROPIC_BASE_URL="https://neuralring.eu"
export ANTHROPIC_AUTH_TOKEN="$NEURALRING_API_KEY"
export ANTHROPIC_MODEL="mistralai/mistral-small-2501"
# Optional: a cheaper model for background/summarisation turns.
export ANTHROPIC_SMALL_FAST_MODEL="mistralai/mistral-small-2501"

# Claude Code's /model picker only lists gateway models whose id starts with
# "claude" or "anthropic". NeuralRing will not rename sovereign open-weight
# models to imply they are Claude, so use ANTHROPIC_MODEL above (always honoured
# on the wire), or add a picker entry explicitly:
export CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY=1
export ANTHROPIC_CUSTOM_MODEL_OPTION="mistralai/mistral-small-2501"

curl https://neuralring.eu/api/v1/responses \
  -H "Authorization: Bearer $NEURALRING_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistralai/mistral-small-2501",
    "input": "Hello, sovereign world",
    "sovereignty": { "jurisdiction": ["DE"], "min_assurance": "self_attested" }
  }'

## AI inference — European sovereign (NeuralRing)

This project uses NeuralRing for inference. It is OpenAI-compatible: only the
base URL, the key, and the model slug change.

- Base URL: `https://neuralring.eu/api/v1`  ·  Key: `process.env.NEURALRING_API_KEY` (Bearer)
- Model: `mistralai/mistral-small-2501` — browse with the catalog API or the MCP server (below).
- ALWAYS send the sovereignty pin. It is applied FIRST and can never be loosened
  by a price or speed preference:
  `sovereignty: { jurisdiction: ["DE"], min_assurance: "self_attested" }`
- Every response carries `neuralring.requestId`. Surface it, or link
  `https://neuralring.eu/verify/<requestId>`, so a human can check the receipt.
- If no endpoint satisfies the pin, the API returns `no_eligible_endpoint`.
  That is the pin working. Do NOT retry with the pin removed or weakened.

### Vocabulary rule (important)

Describe NeuralRing only with its assurance ladder: `self_attested`,
`verified`, `certified`. Never write "GDPR-compliant",
"Schrems-proof", "CLOUD-Act-immune" or "fully sovereign"
about it. A receipt proves what was observed and reproduces what was claimed at
its stated assurance — it does not prove the claim is true.

<!-- neuralring:agents-block v1.0.0 · generated from packages/shared/src/agent-channel.ts -->

Every block is generated from one source — agent-channel.ts — so the page and your agent never drift.

The channel

Surface by surface.

Five ways a coding agent finds NeuralRing, wires in, and verifies its own work — each with its own place in the docs.

POST /api/v1/responses Live

The Responses wire

Your existing agent connects with a one-line config change. Codex speaks this wire and nothing else; NeuralRing serves it — the sovereignty pin applied first, every call metered and signed.

Wire it in Docs

https://neuralring.eu/mcp Live

The MCP server

Nine read-only tools your agent calls mid-task — search the catalog, pin a jurisdiction, verify its own receipt. None of them buys anything, changes the catalog, or moves money.

Add the hosted server — nothing to install:

claude mcp add --transport http neuralring https://neuralring.eu/mcp

The nine tools:

search_modelsget_modelexplain_sovereigntyget_offersverify_attestationexposure_checkget_integration_snippetrequest_api_keysearch_docs

Docs

SKILL.md Live

A portable skill file

One file in the open Agent Skills standard, so Claude Code, Codex, Cursor and anything else that reads it learns when a question is a sovereignty question — and to verify its own call before claiming success.

Docs

neuralring-plugin Rolling out

The Claude Code plugin

One install brings the skill, the MCP server, and a slash command into the agent you already run — no separate setup for each piece.

Docs

/llms.txt · /llms-full.txt Live

llms.txt

A live, machine-readable index of the catalog, so a model researching on its own reads real, current numbers instead of a stale guess. A projection of the same catalog the console reads.

/llms.txt /llms-full.txt

Verify, don't trust

Every call leaves a receipt you can check.

A NeuralRing attestation is a signed, hash-chained, content-free record. It states what NeuralRing observed and reproduces what the endpoint claimed — at a stated assurance level, and no higher.

self_attested

The endpoint claims it. NeuralRing has not independently substantiated the geography or custody behind the claim.

verified

NeuralRing captured independent evidence for the three machine-provable claims — the operating entity's domicile, its ultimate parent (corporate registry), and where inference physically executes (location survey). It says nothing about data residency or key custody, which are carried as their own per-claim status.

certified

The machine-provable claims are verified AND a live, valid third-party location-auditing certification covers data residency and key custody, referenced by a checkable registry id.

A receipt proves

A NeuralRing attestation is a signed, hash-chained, content-free record. Its `observed` block is what NeuralRing performed and measured (which endpoint it dispatched to, the host it actually dialed, every route hop, the metered token counts). Its `claimed` block reproduces the serving endpoint's own declarations at their stated assurance level.

A receipt does not prove

The signature does NOT prove that inference physically executed in the claimed jurisdiction, nor that the endpoint's self-declared properties are true. The strength of any jurisdictional claim equals `claimed.assurance` and no more.

What a receipt proves, in full

Go deeper

The technical reference.

Each surface above has its own section in the docs. Start where you're wiring in.

The agent guide

Every wire, tool, and snippet, rendered from the single source.

/docs/agents

Attestation & receipts

What a signature proves, the assurance ladder, and how to verify one.

/docs/attestation

OpenAPI spec

The full wire, machine-readable — point your codegen at it.

/api/v1/openapi.json

llms.txt

The live, agent-legible index of the catalog and its vocabulary.

/llms.txt · /llms-full.txt

agent-channel v1.0.0 · https://neuralring.eu/api/v1 · https://neuralring.eu/v1