LLM Gateway vs TensorZero
LLM Gateway
llmgateway.io
TensorZero
tensorzero.comSide by side
| Figure | LLM Gateway | TensorZero |
|---|---|---|
| Price from | Free | Free |
| Free plan | Yes | Yes |
| Free trial | Yes | — |
| Open source | Yes | — |
| Web | Yes | Yes |
| Linux | Yes | Yes |
| API | Yes | Yes |
| Self-hosted | Yes | Yes |
| Plans published | 2 | 1 |
| Spec score · LLM gateways | 9.1 (#3) | 8.6 (#7) |
| Spec score · LLM Gateway Software | 8.7 (#6) | 8.0 (#13) |
| Integrations | Published integration guides cover Claude Code, Cursor, Cline, n8n, VS Code, Codex CLI, OpenCode, Continue, Kilo Code, Kimi Code and other coding tools. | Documented providers include Anthropic, AWS Bedrock, Azure, Google, OpenAI, OpenRouter, Together AI, vLLM, and OpenAI-compatible endpoints. |
| Observability | Dashboards expose requests, tokens, latency, errors, cache status and spending broken down by model, provider, project and API key. | TensorZero can store inference and feedback data for observability, optimization, evaluations, and experimentation in Postgres or ClickHouse. |
| Product | LLM Gateway provides one OpenAI-compatible API to route requests across 250+ models from 40+ providers. | TensorZero is an open-source platform that combines an LLM gateway, observability, evaluation, optimization, and experimentation. |
| Security | The company states that it is SOC 2 Type II audited and GDPR-aligned, and describes TLS in transit and AES-256 encryption at rest. | TensorZero's privacy policy says it uses physical, technical, organizational, and administrative security measures to protect personal data. |
| Support | The Free plan lists Discord Community support, while Enterprise includes 24/7 priority support through a dedicated Slack or Discord channel. | The project README directs users to Slack or Discord and offers work teams a Slack or Teams channel by email. |
Plans
In their own words
LLM Gateway
LLM Gateway provides one OpenAI-compatible API for routing requests among more than 250 models from over 40 providers. If a provider errors, times out, or rate-limits a request, it can retry up to twice with another provider for the same model. Its dashboard tracks request, token, spend, latency, error, and cache metrics, with model and provider breakdowns. Teams can use their own provider API keys without a platform fee and pay providers directly. The open-source core is licensed under AGPLv3 and can be self-hosted for free; enterprise features require a license. The Free plan is free, billed free forever, and has a 5% fee on credit usage. It includes 200+ models across 40+ providers, activity logs and export, and 20 requests per minute on free models; paid models are not rate limited. Enterprise pricing is custom and not listed. It includes custom routing, unlimited data retention, a 99.9% uptime SLA, and 24/7 priority support. Published integration guides cover tools including Claude Code, Cursor, Codex CLI, Cline, n8n, and VS Code.
TensorZero
TensorZero is a free, open-source stack that brings together an LLM gateway, observability, optimization, evaluation, and experimentation. Its gateway provides a unified API for model providers and supports routing, retries, fallbacks, prompt templates, schemas, multi-step workflows, and traffic routing for A/B tests. Documented integrations include Anthropic, AWS Bedrock, Azure, Google, OpenAI, OpenRouter, Together AI, vLLM, xAI, and OpenAI-compatible endpoints. Inference and feedback data go to the database you configure, where they can be queried with SQL; feedback can be tied to individual inferences or episodes. The gateway and UI can run through Docker, Docker Compose, Kubernetes and Helm, or a source build. API key authentication is available for gateway endpoints other than status and health checks. The project is available on GitHub, but is no longer maintained; its repository is archived and read-only. TensorZero is self-hosted and licensed under Apache 2.0. It collects pseudonymous aggregate usage analytics, excluding application data and inference inputs and outputs, with an option to disable collection.