- Home
- LLM gateways
- LLM Gateway
LLM Gateway
What it is
LLM Gateway provides one OpenAI-compatible API for routing requests across 250+ models from 40+ providers. If a provider errors, times out, or rate-limits a request, the gateway can retry up to twice with another provider for the same model. Its dashboard tracks requests, tokens, spend, latency, errors, and cache metrics, with breakdowns by model and provider. Users can bring their own provider API keys without a platform fee and pay providers directly. The core gateway is open source under AGPLv3 and can be self-hosted for free; enterprise features require a license. Published integration guides cover tools including Claude Code, Cursor, Codex CLI, Cline, Continue CLI, GitHub Copilot, n8n, OpenCode, and VS Code. The Free plan is free forever, lists a 5% fee on credit usage, and includes 200+ models across 40+ providers. It allows 20 requests per minute on free models; paid models are not rate limited. Enterprise includes custom routing, unlimited data retention, a 99.9% uptime SLA, and 24/7 priority support, with custom pricing.
For and against
For
- One API routes requests across 250+ models.
- Can retry with another provider up to twice.
- Dashboard breaks down usage by model and provider.
- Core gateway can be self-hosted for free.
- BYOK has no platform fee.
Against
- Free plan allows 20 requests per minute on free models.
- Free plan lists a 5% fee on credit usage.
- Enterprise pricing is custom and not listed.
The verdict
LLM Gateway combines provider routing, fallback retries, and request analytics behind one compatible API. Teams should note the Free plan’s listed credit fee and request limit, and that enterprise pricing is custom.
Who it suits
LLM Gateway suits product and platform teams routing model requests across providers. Its self-hosted option and enterprise features also address organizations deploying LLM infrastructure across teams.
Plans and prices
Free
200+ models · 40+ providers · free models rate-limited to 20 requests/minute · metadata-only logs
Read 15 Oct 2026 · llmgateway.io
Enterprise
30-day trial · all models · custom routing · full payload retention for 30 days
Read 15 Oct 2026 · llmgateway.io
On the record
- Analytics
- The dashboard tracks request, token, spend, latency, error, and cache metrics, with breakdowns by model and provider.llmgateway.io · read 7 Oct 2026
- API surface
- Documented endpoints include chat completions, responses, Anthropic messages, embeddings, images, speech, transcription, moderation, search, video and realtime operations.docs.llmgateway.io · read 15 Oct 2026
- Automatic routing
- The gateway selects providers using availability, latency, throughput, price and cache-related signals, with configurable routing behavior.docs.llmgateway.io · read 15 Oct 2026
- BYOK
- Customers may store their own provider keys and use routing and analytics without a platform fee.llmgateway.io · read 15 Oct 2026
- Caching
- Projects can enable response caching to reduce repeated-request latency and token expense.docs.llmgateway.io · read 15 Oct 2026
- Company size
- LinkedIn classifies llm-gateway as a self-employed information-services organization with 2–10 employees.linkedin.com · read 15 Oct 2026
- Credit pricing
- Pay-as-you-go credits use provider list token rates plus a flat 5% fee when credits are purchased; international cards add 1.5%.llmgateway.io · read 15 Oct 2026
- Data retention
- Pricing lists metadata retention as free and full payload retention at $0.01 per 1 million tokens; the Free plan includes activity logs and export.llmgateway.io · read 7 Oct 2026
- Documentation depth
- The documentation includes feature guides, dashboard knowledge-base articles, integration material, self-hosting sections, developer tools and an API reference.docs.llmgateway.io · read 15 Oct 2026
- Enterprise identity
- Enterprise SSO supports SAML 2.0 and OIDC with SCIM provisioning and group-based role mapping.llmgateway.io · read 7 Oct 2026
- Enterprise limits
- Enterprise plans include unlimited seats and unlimited data retention, according to the enterprise page.llmgateway.io · read 7 Oct 2026
- Enterprise pricing
- Enterprise pricing is contact-based and offers volume discounts, custom limits and contractual service levels.llmgateway.io · read 15 Oct 2026
- Failover
- When a provider errors, times out or rate-limits, the gateway can retry on another provider for the same model, up to two times.llmgateway.io · read 15 Oct 2026
- Guardrails
- Enterprise guardrails can detect prompt injection, jailbreak attempts, personally identifying information, payment-card data and secrets, with block, redact or warn actions.llmgateway.io · read 15 Oct 2026
- Identity controls
- Enterprise supports SAML 2.0 or OIDC single sign-on, SCIM provisioning and organization/project roles.llmgateway.io · read 15 Oct 2026
- Independent assessment: API caveat
- GatewayScore reports that latency-based routing applies only to streaming requests and that non-auto tool choices may be downgraded when no mapping supports them.gatewayscore.com · read 15 Oct 2026
- Independent assessment: compliance gap
- GatewayScore says SOC 2 is published but found no unconditional zero-retention guarantee or HIPAA business-associate agreement.gatewayscore.com · read 15 Oct 2026
- Independent assessment: deployment documentation
- GatewayScore found hosted and self-host deployment documented, while VPC, on-premises and air-gapped deployment were not documented on the reviewed pages.gatewayscore.com · read 15 Oct 2026
