# ZeroGPU

> ZeroGPU is the compute efficiency layer for AI inference. It runs repeatable, high-volume tasks - classification, extraction, PII redaction, moderation, summarization, routing - on specialized small and nano language models across an edge-powered network, faster and cheaper than centralized GPUs, through one OpenAI-compatible API.

## Docs

- [Introduction](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/index.md): Run high-volume AI tasks on specialized small and nano language models across an edge-powered network - faster and cheaper, through one OpenAI-compatible API.
- [How ZeroGPU works](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/how-zerogpu-works.md): ZeroGPU turns idle compute into one programmable inference layer - specialized small models, an edge-powered network, and routing handled for you.
- [Quickstart](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/quickstart.md): First API call in minutes.
- [Model Catalog](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/model-catalog.md): Browse ZeroGPU models, compare pricing, and pick the model ID to call.
- [Open Weight](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/open-weight.md): The open-weight frontier-class models served on ZeroGPU, with pricing and context windows.
- [Text Generation](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/text-generation.md): Edge generative models for conversation and reasoning.
- [Text Classification](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/text-classification.md): Edge models that map text to labels, IAB taxonomy, and audience signals.
- [Moderation](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/moderation.md): OpenAI-compatible content moderation, benchmarked against omni-moderation-latest.
- [Data Extraction](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/pii.md): Multilingual PII detection, redaction, and schema-driven extraction models.
- [Ad Tech](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/ad-tech.md): Edge IAB classifiers that map content to taxonomy, audience, and intent signals for programmatic advertising.
- [Overview](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/index.md): Run large workloads of AI inference asynchronously, wire-compatible with `/v1/batches` and `/v1/files`.
- [Quickstart](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/getting-started.md): First batch in under 10 minutes, auth, upload, create, poll, download.
- [Objects & lifecycle](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/objects.md): The File and Batch objects, the batch status lifecycle, and how the two APIs fit together.
- [JSONL Format](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/jsonl-format.md): Mandatory line schemas for input, output, and error JSONL files.
- [Supported Endpoints](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/supported-endpoints.md): The only endpoint a batch line can target. Same request body, same response body, only the routing is asynchronous.
- [Errors](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/batch/errors.md): Every HTTP status, every validation message, every JSONL error code, and how to recover.
- [Production patterns](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/production-patterns.md): What to get right when you move ZeroGPU from a first call into production.
- [Platform](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/docs/platform.md): Everything for managing your account, credentials, and usage in one place - your organization and projects, API keys, authentication, usage and logs, and billing.
- [Overview](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/index.md): Plug ZeroGPU's nano language models into the tools and runtimes you already use.
- [Claude Code Plugin](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/claude-code-plugin.md): Route Claude Code subagents and slash commands through ZeroGPU's nano language models.
- [Claude Skill](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/claude-skill.md): Upload the ZeroGPU Skill to Claude (Desktop or Web) so Claude follows ZeroGPU best practices when you ask it to classify, extract, redact, or summarize content.
- [Hermes Agent](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/hermes.md): Run Hermes Agent on ZeroGPU: set it as the model provider and connect the ZeroGPU MCP server for thirteen ready-made tools.
- [LangChain](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/langchain.md): Offload classification, extraction, PII redaction, and chat from LangChain agents to ZeroGPU's nano language models with eleven ready-made tools.
- [MCP Server](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/mcp.md): Connect any MCP client to ZeroGPU's nano language models over the Model Context Protocol.
- [OpenClaw](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/openclaw.md): Route your OpenClaw agent's routine tasks through ZeroGPU's nano language models with the ZeroGPU Router plugin.
- [Go](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/sdk/go.md): Integrate ZeroGPU into your Go application.
- [JavaScript](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/sdk/javascript.md): Integrate ZeroGPU into your JavaScript or Node.js application.
- [Python](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/sdk/python.md): Integrate ZeroGPU into your Python application.
- [Ruby](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/sdk/ruby.md): Integrate ZeroGPU into your Ruby application.
- [Rust](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/integrations/sdk/rust.md): Integrate ZeroGPU into your Rust application.
- [Responses](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/responses.md): Send input to an AI model and receive a response.
- [Chat completions](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/chat-completions.md): OpenAI-style chat inference for models that use this route.
- [Moderations](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/moderations.md): Classify text against OpenAI's safety categories.
- [gpt-oss-120b](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/gpt-oss-120b.md): Model details for gpt-oss-120b. Function calling, batch tasks, and reasoning with a 131K context window.
- [qwen3-30b-a3b-fp8](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/qwen3-30b-a3b-fp8.md): Model details for qwen3-30b-a3b-fp8. Multilingual, streaming, batch tasks, reasoning, and function calling with a 32K context window.
- [glm-5.2](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/glm-5-2.md): Model details for glm-5.2. Reasoning, function calling, and batch tasks with a 256K-token context window.
- [deepseek-v4-flash](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/deepseek-v4-flash.md): Model details for deepseek-v4-flash. Reasoning, coding, and agentic workflows with a 1M-token context window.
- [deberta-v3-small](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/deberta-v3-small.md): Model details for deberta-v3-small.
- [gliner-multi-pii-v1](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/gliner-multi-pii-v1.md): Model details for gliner-multi-pii-v1.
- [gliner2-base-v1](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/gliner2-base-v1.md): Model details for gliner2-base-v1.
- [LFM2.5-1.2B-Instruct](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/lfm2-5-1-2b-instruct.md): Model details for LFM2.5-1.2B-Instruct.
- [LFM2.5-1.2B-Thinking](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/lfm2-5-1-2b-thinking.md): Model details for LFM2.5-1.2B-Thinking.
- [llama-3.1-8b-instruct-fast](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/llama-3-1-8b-instruct-fast.md): Model details for llama-3.1-8b-instruct-fast.
- [zlm-v1-iab-classify-edge](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/zlm-v1-iab-classify-edge.md): Model details for zlm-v1-iab-classify-edge. Supports 50+ languages.
- [zlm-v2-iab-classify-edge-enriched](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/zlm-v2-iab-classify-edge-enriched.md): Model details for zlm-v2-iab-classify-edge-enriched. Supports 50+ languages.
- [zlm-v1-iab-domain-classifier](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/zlm-v1-iab-domain-classifier.md): Model details for zlm-v1-iab-domain-classifier.
- [zlm-v1-moderation-edge](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/models/zlm-v1-moderation-edge.md): Model details for zlm-v1-moderation-edge. OpenAI-compatible moderation, benchmarked against omni-moderation-latest.
- [Responses (ZeroClick)](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/zeroclick-responses.md): OpenAI-style Responses inference authenticated by a ZeroClick HMAC signature instead of an API key.
- [Chat completions (ZeroClick)](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/zeroclick-chat-completions.md): OpenAI-style chat inference authenticated by a ZeroClick HMAC signature instead of an API key.
- [Upload file](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/upload-file.md): Interactive playground for POST /v1/files (multipart JSONL upload).
- [List files](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/list-files.md): Interactive playground for GET /v1/files.
- [Retrieve file](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/retrieve-file.md): Interactive playground for GET /v1/files/{file_id}.
- [Download file](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/download-file.md): Interactive playground for GET /v1/files/{file_id}/content.
- [Delete file](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/delete-file.md): Interactive playground for DELETE /v1/files/{file_id}.
- [Create batch](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/create-batch.md): Interactive playground for POST /v1/batches.
- [Retrieve batch](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/retrieve-batch.md): Interactive playground for GET /v1/batches/{batch_id}.
- [List batches](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/list-batches.md): Interactive playground for GET /v1/batches.
- [Cancel batch](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/cancel-batch.md): Interactive playground for POST /v1/batches/{batch_id}/cancel.
- [Overview](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/index.md): Ready-to-run recipes for the high-volume tasks ZeroGPU's nano models are built for.
- [🎓 Resume & profile extraction](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/data-extraction-tutorial.md): Step-by-step guide with example dataset. Text-only extraction, no OCR.
- [🏷️ Tag Customer Reviews Overnight with the ZeroGPU Batch API](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/batch-review-tagging.md): Tag a CSV of customer reviews with sentiment and topics in one overnight job using the ZeroGPU Batch API and LFM2.5-1.2B-Instruct, then recover any failed rows by custom_id.
- [🔐 Sanitize a CSV of Customer Feedback with the ZeroGPU Router Plugin](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/claude-code-csv-sanitizer.md): Use the zerogpu-router plugin so Claude Code can redact names, emails, and phone numbers from a raw CSV export in one prompt, returning a clean copy plus a PII audit log.
- [🦜 Screen Resumes with LangChain and ZeroGPU](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/langchain-resume-screening.md)
- [🎯 Turn Raw Content into Targeted Copy with the IAB Classifier and Claude](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/claude-iab-targeted-copy.md): Connect the ZeroGPU MCP server in Claude Desktop so Claude can call the zlm-v1-iab-classify-edge model, then turn the returned IAB categories, audience segments, and confidence scores into three targeted outputs: an ad brief, a newsletter blurb, and a content pitch.
- [🎫 Route Support Tickets with Claude and ZeroGPU's Thinking Model](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/claude-support-ticket-routing.md): Use the ZeroGPU Skill in Claude to triage incoming support tickets. Claude builds an exact /v1/responses call to the LFM2.5-1.2B-Thinking model, which returns a team and priority for routing.
- [🦞 Build a PII-Safe Inbox Summarizer with Your OpenClaw Assistant](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/openclaw-inbox-summarizer.md): Forward messy messages to your OpenClaw assistant and have it redact PII, summarize the thread, and classify topic and urgency, offloading every cheap step to ZeroGPU's nano models.
- [🏛️ Raise Your Seed Round with a Hermes Agent](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/cookbook/hermes-investor-outreach.md): Point Hermes Agent at ZeroGPU, then let it research investors, rank them against your pitch, write a personalized email for each, and save them all as Gmail drafts for a fraction of a frontier-model run.

## OpenAPI Specs

- [zerogpu.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/zerogpu.openapi.json)
- [zeroclick.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/zeroclick.openapi.json)
- [zlm-v2-iab-classify-edge-enriched.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/zlm-v2-iab-classify-edge-enriched.openapi.json)
- [zlm-v1-moderation-edge.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/zlm-v1-moderation-edge.openapi.json)
- [zlm-v1-iab-domain-classifier.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/zlm-v1-iab-domain-classifier.openapi.json)
- [zlm-v1-iab-classify-edge.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/zlm-v1-iab-classify-edge.openapi.json)
- [qwen3-30b-a3b-fp8.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/qwen3-30b-a3b-fp8.openapi.json)
- [llama-3_1-8b-instruct-fast.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/llama-3_1-8b-instruct-fast.openapi.json)
- [gpt-oss-120b.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/gpt-oss-120b.openapi.json)
- [glm-5_2.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/glm-5_2.openapi.json)
- [gliner2-base-v1.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/gliner2-base-v1.openapi.json)
- [gliner-multi-pii-v1.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/gliner-multi-pii-v1.openapi.json)
- [deepseek-v4-flash.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/deepseek-v4-flash.openapi.json)
- [deberta-v3-small.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/deberta-v3-small.openapi.json)
- [LFM2_5-1_2B-Thinking.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/LFM2_5-1_2B-Thinking.openapi.json)
- [LFM2_5-1_2B-Instruct.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/playgrounds/LFM2_5-1_2B-Instruct.openapi.json)
- [batch.openapi](https://zerogpu-claude-friendly-johnson-fww5io.mintlify.site/api-reference/openapi/batch.openapi.json)

## Optional

- [Website](https://zerogpu.ai)
- [Platform](https://platform.zerogpu.ai/)
- [llms.txt](https://docs.zerogpu.ai/llms.txt)
- [llms-full.txt](https://docs.zerogpu.ai/llms-full.txt)