# Overshoot > Overshoot is a real-time video understanding API: ingest live frames over WebRTC and query them with vision-language models through an OpenAI-compatible interface. ## Docs - [Welcome to Overshoot](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/intro.md) - [Quickstart](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/quickstart.md): Get started in four quick steps - [The Stream](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/the-stream.md): A leased session that holds your live video feed and makes it addressable from any chat completion. - [Chat Completion](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/chat-completion.md): Use an OpenAI-compatible request to interact with your stream. Reference any moment by URL: a single frame, a window, or several windows. - [Models](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/models.md): Vision Language Models on Overshoot - [Glossary](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/glossary.md): Canonical definitions for Overshoot terms used across the docs and API. - [Best practices](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/best-practices.md): Patterns for keeping latency low, costs predictable, and streams alive. - [Introduction](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/introduction.md): Use the Overshoot REST API to ingest live video, query vision-language models, and manage stream lifecycle. - [Authentication](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/authentication.md): Bearer API keys for every public HTTP request, except the open model and pricing endpoints. - [Core flow](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/core-flow.md): The end-to-end lifecycle: create a stream, publish frames, query frames with chat completions. - [Stream lifecycle](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/stream-lifecycle.md): How streams transition between active and ended, and how frame availability is exposed. - [Stream media URLs](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/stream-media-urls.md): The ovs:// URL grammar for referencing frames and segments inside chat completions. - [Chat completions](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/chat-completions.md): OpenAI-compatible chat completions. - [Create stream](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/create-stream.md): Creates a live stream and returns a publish token. The current request body is empty. The retained frame history is fixed at 600 seconds. - [Get stream](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/get-stream.md): Returns current stream status and frame telemetry. - [Keepalive](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/keepalive-stream.md): Renews the stream lease and returns a fresh publish token. Clients should renew before the 5-minute lease expires, for example every 4 minutes, and pass the returned token to the LiveKit SDK with `room.updateToken(...)`. - [Delete stream](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/delete-stream.md): End a stream and release its resources. Idempotent on already-deleted streams within the lookup window. - [List models](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/list-models.md): List model IDs and availability in an OpenAI-compatible shape. No auth required. Use the `id` value as the `model` field in inference requests. - [List pricing](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/list-pricing.md): List public model prices. No auth required. Money fields use **integer microcents**: `1 cent = 1,000,000 microcents`. - [Get model pricing](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/get-pricing.md): Return pricing for one model. The `model` path may contain slashes. - [Get prepaid balance](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/get-balance.md): Return the authenticated user's prepaid balance. `balance_cents` is for display; `balance_microcents` is the exact billing unit. Inference can return `402` when billing denies a request. - [Create checkout session](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/create-checkout.md): Create a checkout session for buying prepaid credits. `amount_cents` must be at least `100`. - [LiveKit client flow](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/livekit-client-flow.md): Browser example showing how to create a stream, publish via LiveKit, and renew the lease. - [Region behavior](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/regions.md): How regional routing works and how to skip the redirect/retry step. - [Prompt cache and threads](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/prompt-cache.md): Reuse cached prompt prefixes across related chat completion requests. - [Error shapes](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/errors.md): Status codes, error codes, and the JSON envelopes used across stream and chat endpoints. - [Limits and retention](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/limits.md): Frame retention window and other API limits. ## OpenAPI Specs - [openapi](https://overshoot-fix-llm-text-renew-stream-lease.mintlify.site/api-reference/openapi.yaml)