Overview

Gateway overview

Requests / min

1,240

p95 latency

380 ms

Spend today

€84.20

Error rate

0.4 %

Traffic · last 24h

Models

Provider	Model	Status	Routing weight	Avg latency	Cost / 1k tokens
Anthropic	claude-sonnet-5	Active	45%	410 ms	€0.009
OpenAI	gpt-4o	Active	30%	520 ms	€0.011
Mistral	mistral-large	Active	20%	240 ms	€0.004
Meta	llama-3.3-70b (self-hosted)	Degraded	5%	180 ms	€0.000

Recent requests

Time	Model	Tokens in / out	Status	Latency
11:49:02	claude-sonnet-5	1,204 / 380	200	412 ms
11:48:57	mistral-large	890 / 156	200	233 ms
11:48:51	gpt-4o	2,010 / 640	200	540 ms
11:48:44	claude-sonnet-5	540 / 0	429	rate-limited
11:48:39	llama-3.3-70b	1,100 / 420	200	176 ms

Models

Model catalog

Provider	Model	Status	Routing weight	Avg latency	Cost / 1k tokens
Anthropic	claude-sonnet-5	Active	45%	410 ms	€0.009
OpenAI	gpt-4o	Active	30%	520 ms	€0.011
Mistral	mistral-large	Active	20%	240 ms	€0.004
Meta	llama-3.3-70b (self-hosted)	Degraded	5%	180 ms	€0.000

Routes

Routing rules

chat-default

Primary conversational route

→ claude-sonnet-5

summarization

Long-document summarization

→ gpt-4o

bulk-classify

High-volume, cost-sensitive route

→ mistral-large

fallback

Used when primary route is unavailable

→ llama-3.3-70b

Policies

Governance policies

Rate limit per key

60 requests / min

PII redaction

Strip emails and card numbers before logging

Allowed models allowlist

4 models approved

EnterprisePer-team quotas

Unlock →

Team	Monthly quota	Used
Product Engineering	2.0M tokens	64%
Data Science	1.2M tokens	41%
Customer Support	500k tokens	88%

EnterprisePII redaction rules

Unlock →

Rule	Pattern	Action
Email address	RFC 5322	Redact
Card number	Luhn-validated	Redact
National ID	Custom regex	Flag & log

EnterpriseSingle sign-on (SSO)

Unlock →

SAML / OIDC provider

Okta, Entra ID or any OIDC-compliant IdP

Requests

Request log

Time	Model	Tokens in / out	Status	Latency
11:49:02	claude-sonnet-5	1,204 / 380	200	412 ms
11:48:57	mistral-large	890 / 156	200	233 ms
11:48:51	gpt-4o	2,010 / 640	200	540 ms
11:48:44	claude-sonnet-5	540 / 0	429	rate-limited
11:48:39	llama-3.3-70b	1,100 / 420	200	176 ms

Usage & cost

Analytics

EnterpriseUsage & cost analytics

Unlock →

Model	Requests	Tokens	Cost today
claude-sonnet-5	14,204	7.2M	€38.10
gpt-4o	9,340	5.1M	€27.85
mistral-large	6,012	2.6M	€14.90
llama-3.3-70b	1,540	0.7M	€3.35

Entitlements

License

Self-hosted · Community

License key

Running Community, self-hosted — Apache 2.0. Enter a license key to unlock per-team quotas, granular PII rules, SSO and usage analytics.

Talk to engineering

Self-hosted · Enterprise

License key

Signed entitlement · valid until 2027-01-31

All Enterprise capabilities are active on this deployment.

Community · Hosted€0 / mo

Included requests50,000 / mo

Next invoice—

Hosted by VEV, Community feature set. Upgrade to Enterprise for quotas, SSO and analytics.

Talk to engineering

Enterprise · Hosted€900 / mo

Included requests2,000,000 / mo

Overage rate€0.0006 / request

Next invoice2026-09-01 · €1,142.30

Managed by VEV. All Enterprise capabilities are active.