# LLM Stats

> LLM Stats is powered by ZeroEval, the open-source evaluation library built by the LLM Stats team

## Docs

- [Get Model](https://docs.llm-stats.com/api-reference/get-model.md)
- [Get Rankings](https://docs.llm-stats.com/api-reference/get-rankings.md)
- [Introduction](https://docs.llm-stats.com/api-reference/introduction.md): Programmatic access to benchmark scores, model rankings, pricing, and metadata across every model we track.
- [List Benchmarks](https://docs.llm-stats.com/api-reference/list-benchmarks.md)
- [List Models](https://docs.llm-stats.com/api-reference/list-models.md)
- [List Scores](https://docs.llm-stats.com/api-reference/list-scores.md)
- [List Updates](https://docs.llm-stats.com/api-reference/list-updates.md)
- [MCP Server](https://docs.llm-stats.com/api-reference/mcp.md): Query the LLM Stats catalog and call the gateway from any MCP-compatible client.
- [Chat completions](https://docs.llm-stats.com/gateway/chat-completions.md): OpenAI-compatible chat completions with automatic provider routing.
- [Errors](https://docs.llm-stats.com/gateway/errors.md): One error envelope for every gateway endpoint.
- [Image & video generations](https://docs.llm-stats.com/gateway/generations.md): One unified endpoint for every image and video model, with built-in long-polling.
- [Introduction](https://docs.llm-stats.com/gateway/introduction.md): One endpoint per modality, every provider behind it.
- [Speech-to-text](https://docs.llm-stats.com/gateway/speech-to-text.md): Transcribe audio files with one HTTP call.
- [Text-to-speech](https://docs.llm-stats.com/gateway/text-to-speech.md): Synthesize audio from text with one HTTP call.
- [ZeroEval Python SDK](https://docs.llm-stats.com/index.md): LLM Stats is powered by ZeroEval, our in-house evaluation library
- [Using the CLI](https://docs.llm-stats.com/python-sdk/cli/using-the-cli.md): Inspect datasets and evals quickly with the ZeroEval CLI
- [Creating Datasets](https://docs.llm-stats.com/python-sdk/datasets/creating.md): Create benchmark datasets from the UI, git, or the Python SDK
- [Loading and Inspecting Datasets](https://docs.llm-stats.com/python-sdk/datasets/loading.md): Pull datasets, iterate rows, and work with slices
- [Multimodal Datasets](https://docs.llm-stats.com/python-sdk/datasets/multimodal.md): Attach images, audio, video, and media URLs to dataset rows
- [Uploading Data](https://docs.llm-stats.com/python-sdk/datasets/uploading-data.md): Add benchmark data via git, CSV import, or the browser editor
- [Versioning and Subsets](https://docs.llm-stats.com/python-sdk/datasets/versioning-and-subsets.md): Control dataset versions and subset selection for reproducible evals
- [Common Failures](https://docs.llm-stats.com/python-sdk/evals/common-failures.md): Debug frequent dataset and eval errors quickly
- [Repetitions and Resume](https://docs.llm-stats.com/python-sdk/evals/repetitions-and-resume.md): Repeat runs for uncertainty and resume interrupted executions safely
- [Running Evals](https://docs.llm-stats.com/python-sdk/evals/running-evals.md): Execute tasks on datasets with concurrency, retries, and checkpointing
- [Scoring and Metrics](https://docs.llm-stats.com/python-sdk/evals/scoring-and-metrics.md): Use row, column, and run-level evaluations with explicit column mapping
- [Signals and Traces](https://docs.llm-stats.com/python-sdk/evals/signals-and-traces.md): Emit runtime signals during task execution and inspect them through traces
- [Multimodal Eval End-to-End](https://docs.llm-stats.com/python-sdk/examples/multimodal-eval-end-to-end.md): Run a multimodal evaluation with image inputs and row-level scoring
- [Text Eval End-to-End](https://docs.llm-stats.com/python-sdk/examples/text-eval-end-to-end.md): A complete text evaluation workflow with row and column metrics
- [Installation](https://docs.llm-stats.com/python-sdk/installation.md): Install and initialize the ZeroEval Python SDK
- [Python SDK Quickstart](https://docs.llm-stats.com/python-sdk/quickstart.md): End-to-end quickstart for dataset + eval + metrics
- [Quickstart](https://docs.llm-stats.com/quickstart.md): Run your first dataset evaluation with the ZeroEval Python SDK

## OpenAPI Specs

- [openapi](https://docs.llm-stats.com/api-reference/openapi.json)