TensorZero is an open-source LLMOps platform that unifies:
Gateway: access every LLM provider through a unified API, built for performance (<1ms p99 latency)Observability: store inferences and feedback in your database, available programmatically or in the UIEvaluation: benchmark individual inferences or end-to-end workflows using heuristics, LLM judges, etc.Optimization: collect metrics and human feedback to optimize prompts, models, and inference strategiesExperimentation: ship with confidence with built-in A/B testing, routing, fallbacks, retries, etc.
You can take what you need, adopt incrementally, and complement with other tools. It plays nicely with the OpenAI SDK, OpenTelemetry, and every major LLM provider.
TensorZero is used by companies ranging from frontier AI startups to the Fortune 10 and fuels ~1% of global LLM API spend today.
** Website** ·
·
·
Configuration Reference## tensorzero-demo.mp4
Note
TensorZero Autopilot is an automated AI engineer powered by TensorZero that analyzes LLM observability data, sets up evals, optimizes prompts and models, and runs A/B tests.
It dramatically improves the performance of LLM agents across diverse tasks:
Integrate with TensorZero once and access every major LLM provider.
(API or self-hosted) through a single unified APICall any LLM - Infer with ,tool use,structured outputs (JSON),batch,embeddings,multimodal (images, files), etc.caching - to enforce a structured interface between your application and the LLMsCreate prompt templates and schemas - Satisfy extreme throughput and latency needs, thanks to 🦀 Rust: <1ms p99 latency overhead at 10k+ QPS - with routing, retries, fallbacks, load balancing, granular timeouts, etc.Ensure high availability - andTrack usage and costwith granular scopes (e.g. tags)enforce custom rate limits - to allow clients to access models without sharing provider API keysSet up auth for TensorZero
** Anthropic**,
,
Google AI Studio (Gemini API),
Groq,
TGI,
Together AI, and
vLLM.
xAI (Grok)Need something else? TensorZero also supports ** any OpenAI-compatible API (e.g. Ollama)**.
You can use TensorZero with any OpenAI SDK (Python, Node, Go, etc.) or OpenAI-compatible client.
(one Docker container).Deploy the TensorZero Gateway- Update the
base_url
andmodel
in your OpenAI-compatible client. - Run inference:
from openai import OpenAI
client = OpenAI(base_url="http://localhost:3000/openai/v1", api_key="not-used")
response = client.chat.completions.create(
model="tensorzero::model_name::anthropic::claude-sonnet-4-6",
messages=[
{
"role": "user",
"content": "Share a fun fact about TensorZero.",
}
],
)
See ** Quick Start** for more information.
Zoom in to debug individual API calls, or zoom out to monitor metrics across models and prompts over time — all using the open-source TensorZero UI.
- Store inferences and in your own databasefeedback (metrics, human edits, etc.) - Dive into individual inferences or high-level aggregate patterns using the TensorZero UI or programmatically
for optimization, evaluation, and other workflowsBuild datasets - Replay historical inferences with new prompts, models, inference strategies, etc. #
andExport OpenTelemetry traces (OTLP)to your favorite application observability toolsexport Prometheus metrics - Soon: AI-assisted debugging and root cause analysis; AI-assisted data labeling
Send production metrics and human feedback to easily optimize your prompts, models, and inference strategies — using the UI or programmatically.
- Optimize your models with , RLHF, and other techniquessupervised fine-tuning - Optimize your prompts with automated prompt engineering algorithms like GEPA - Optimize your withinference strategy, best/mixture-of-N sampling, etc.dynamic in-context learning - Enable a feedback loop for your LLMs: a data & learning flywheel turning production data into smarter, faster, and cheaper models
- Soon: synthetic data generation
Compare prompts, models, and inference strategies using evaluations powered by heuristics and LLM judges.
withEvaluate individual inferencesinference evaluationspowered by heuristics or LLM judges (≈ unit tests for LLMs) - withEvaluate end-to-end workflowsworkflow evaluationswith complete flexibility (≈ integration tests for LLMs) - Optimize LLM judges just like any other TensorZero function to align them to human preferences
- Soon: more built-in evaluators; headless evaluations
Evaluation » UI | Evaluation » CLI | |
docker compose run --rm evaluations \
--evaluation-name extract_data \
--dataset-name hard_test_cases \
--variant-name gpt_4o \
--concurrency 5
Run ID: 01961de9-c8a4-7c60-ab8d-15491a9708e4
Number of datapoints: 100
██████████████████████████████████████ 100/100
exact_match: 0.83 ± 0.03 (n=100)
semantic_match: 0.98 ± 0.01 (n=100)
item_count: 7.15 ± 0.39 (n=100)
|
Ship with confidence with built-in A/B testing, routing, fallbacks, retries, etc.
to ship with confidence and identify the best prompts and models for your use cases.Run adaptive A/B tests - Enforce principled experiments in complex workflows, including support for multi-turn LLM systems, sequential testing, and more.
Build with an open-source stack well-suited for prototypes but designed from the ground up to support the most complex LLM applications and deployments.
- Build simple applications or massive deployments with GitOps-friendly orchestration
with built-in escape hatches, programmatic-first usage, direct database access, and moreExtend TensorZero - Integrate with third-party tools: specialized observability and evaluations, model providers, agent orchestration frameworks, etc.
- Iterate quickly by experimenting with prompts interactively using the Playground UI
How is TensorZero different from other LLM frameworks?
- TensorZero enables you to optimize complex LLM applications based on production metrics and human feedback.
- TensorZero supports the needs of industrial-grade LLM applications: low latency, high throughput, type safety, self-hosted, GitOps, customizability, etc.
- TensorZero unifies the entire LLMOps stack, creating compounding benefits. For example, LLM evaluations can be used for fine-tuning models alongside AI judges.
Can I use TensorZero with ___?
Yes. Every major programming language is supported. It plays nicely with the OpenAI SDK, OpenTelemetry, and every major LLM provider.
Is TensorZero production-ready?
Yes. TensorZero is used by companies ranging from frontier AI startups to the Fortune 10 and powers ~1% of the global LLM API spend today.
Here's a case study: Automating Code Changelogs at a Large Bank with LLMs
How much does TensorZero cost?
TensorZero (LLMOps platform) is 100% self-hosted and open-source.
TensorZero Autopilot (automated AI engineer) is a complementary paid product powered by TensorZero.
Who is building TensorZero?
Our technical team includes a former Rust compiler maintainer, machine learning researchers (Stanford, CMU, Oxford, Columbia) with thousands of citations, and the chief product officer of a decacorn startup. We're backed by the same investors as leading open-source projects (e.g. ClickHouse, CockroachDB) and AI labs (e.g. OpenAI, Anthropic). See our ** $7.3M seed round announcement** and
. We're
hiring in NYCHow do I get started?
You can adopt TensorZero incrementally. Our ** Quick Start** goes from a vanilla OpenAI wrapper to a production-ready LLM application with observability and fine-tuning in just 5 minutes.
Start building today. The ** Quick Start** shows it's easy to set up an LLM application with TensorZero.
Questions? Ask us on ** Slack** or
.
DiscordUsing TensorZero at work? Email us at ** hello@tensorzero.com** to set up a Slack or Teams channel with your team (free).
We are working on a series of complete runnable examples illustrating TensorZero's data & learning flywheel.
[Optimizing Data Extraction (NER) with TensorZero]This example shows how to use TensorZero to optimize a data extraction pipeline. We demonstrate techniques like fine-tuning and dynamic in-context learning (DICL). In the end, an optimized GPT-4o Mini model outperforms GPT-4o on this task — at a fraction of the cost and latency — using a small amount of training data.
[Agentic RAG — Multi-Hop Question Answering with LLMs]This example shows how to build a multi-hop retrieval agent using TensorZero. The agent iteratively searches Wikipedia to gather information, and decides when it has enough context to answer a complex question.
[Writing Haikus to Satisfy a Judge with Hidden Preferences]This example fine-tunes GPT-4o Mini to generate haikus tailored to a specific taste. You'll see TensorZero's "data flywheel in a box" in action: better variants leads to better data, and better data leads to better variants. You'll see progress by fine-tuning the LLM multiple times.
[Image Data Extraction — Multimodal (Vision) Fine-tuning]This example shows how to fine-tune multimodal models (VLMs) like GPT-4o to improve their performance on vision-language tasks. Specifically, we'll build a system that categorizes document images (screenshots of computer science research papers).
[Improving LLM Chess Ability with Best-of-N Sampling]This example showcases how best-of-N sampling can significantly enhance an LLM's chess-playing abilities by selecting the most promising moves from multiple generated options.
We write about LLM engineering on the ** TensorZero Blog**. Here are some of our favorite posts:
Bandits in your LLM Gateway: Improve LLM Applications Faster with Adaptive Experimentation (A/B Testing)Is OpenAI's Reinforcement Fine-Tuning (RFT) Worth It?Distillation with Programmatic Data Curation: Smarter LLMs, 5-30x Cheaper InferenceFrom NER to Agents: Does Automated Prompt Engineering Scale to Complex Tasks?