# OmniProvider
> Pay-per-call LLM API for AI agents. A multi-model catalog (OpenAI-style chat completions) and cited briefs, paid per request in USDC over x402 (Base) or MPP (Tempo), plus prepaid packages. No account or API key for pay-per-call. Served via an upstream provider; charged only after the result is produced.

The API host's own llms.txt (https://pay.omnidev.tech/llms.txt) is the contract of record for routes, prices and payment details.

## Model catalog (live data as of 2026-10-17 06:20 UTC; authoritative: https://pay.omnidev.tech/v1/models)
- kimi-k3: input $0.08 / 1M tokens, output $0.24 / 1M tokens, context 1,048,576, max output 8,192, tier open; Live (verified 2026-10-16)
- deepseek-v4.1-flash: input $0.06 / 1M tokens, output $0.06 / 1M tokens, context 5,000,000, max output 8,192, tier open; Live (verified 2026-10-16)
- glm-5.3: input $0.40 / 1M tokens, output $0.20 / 1M tokens, context 5,000,000, max output 8,192, tier open; Live (verified 2026-10-16)
- llama-3.1-8b: input $0.06 / 1M tokens, output $0.12 / 1M tokens, context 16,384, max output 8,192, tier open; Live (verified 2026-10-16)
- gpt-oss-120b: input $0.12 / 1M tokens, output $0.50 / 1M tokens, context 122,880, max output 8,192, tier open; Live (verified 2026-10-16)
- gpt-6-luna: input $0.15 / 1M tokens, output $0.45 / 1M tokens, context 1,000,000, max output 8,192, tier frontier-lite; Live (verified 2026-10-16)
- claude-haiku-5.5: input $0.15 / 1M tokens, output $0.75 / 1M tokens, context 100,000, max output 8,192, tier frontier-lite; Live (verified 2026-10-16)
- gpt-6-sol: input $2.40 / 1M tokens, output $12.00 / 1M tokens, context 1,000,000, max output 8,192, tier frontier-pro; Live (verified 2026-10-16)
- claude-sonnet-5.5: input $2.40 / 1M tokens, output $12.00 / 1M tokens, context 1,000,000, max output 8,192, tier frontier-pro; Live (verified 2026-10-16)
- gemini-3.1-pro: input $2.40 / 1M tokens, output $14.40 / 1M tokens, context 200,000, max output 8,192, tier frontier-pro; Live (verified 2026-10-16)
- grok-4.7: input $2.40 / 1M tokens, output $7.20 / 1M tokens, context 500,000, max output 8,192, tier frontier-pro; Live (verified 2026-10-16)
- gpt-6-astra: input $12.00 / 1M tokens, output $60.00 / 1M tokens, context 1,000,000, max output 8,192, tier frontier-max; Listed; the per-request MPP price cap limits output length on this model, so use a prepaid package when frontier credit opens
- glm-5.3-flash-uncensored: input $0.30 / 1M tokens, output $1.00 / 1M tokens, context 131,072, max output 2,048, tier open-aup; Listed, end-to-end check pending; restricted access (18+), per-call only, see https://provider.omnidev.tech/docs/restricted
- mimo-v2.6-flash-abliterated: input $1.50 / 1M tokens, output $4.50 / 1M tokens, context 131,072, max output 2,048, tier open-aup; Listed, end-to-end check pending; restricted access (18+), per-call only, see https://provider.omnidev.tech/docs/restricted

Endpoint: POST https://pay.omnidev.tech/v1/chat/completions with {"model":"<id>","messages":[...],"max_tokens":256} (MPP / Tempo; prepaid package API keys also accepted). Prices are USD per 1M tokens. See https://provider.omnidev.tech/catalog

## API
- [OpenAI-style model list](https://pay.omnidev.tech/v1/models): live prices, context, rails
- [Prepaid packages](https://pay.omnidev.tech/v1/packages): token packs and frontier credit
- [OpenAPI description](https://pay.omnidev.tech/openapi.json): routes, prices, payment protocols, request/response schemas
- [Token bundles](https://pay.omnidev.tech/v1/bundles): prepaid token bundles (10M / 100M tokens), pay once over MPP, draw down per call, wallet-bound balance
- [x402 discovery](https://pay.omnidev.tech/.well-known/x402): paid resources for x402 clients
- [API llms.txt](https://pay.omnidev.tech/llms.txt): endpoints, prices, nonce header, examples
- [Health](https://pay.omnidev.tech/health): service status
- [Agent card](https://provider.omnidev.tech/.well-known/agent-card.json): skills, routes, payment protocols (also /.well-known/agent.json)
- [MCP discovery](https://provider.omnidev.tech/.well-known/mcp.json): remote MCP server (read-only discovery tools) at https://mcp.omnidev.tech/mcp
- [MCP server card](https://provider.omnidev.tech/.well-known/mcp/server-card.json)
- [AI plugin manifest](https://provider.omnidev.tech/.well-known/ai-plugin.json): points to the OpenAPI description

## Docs
- [Overview and pricing](https://provider.omnidev.tech/): how it works, payment rails, live prices
- [Model catalog](https://provider.omnidev.tech/catalog): every model, price per 1M tokens, context, rails, status
- [Quick start](https://provider.omnidev.tech/docs/quickstart): price check, catalog call over MPP, x402 fixed-price routes, packages
- [API reference](https://provider.omnidev.tech/docs/api): every route, generated from the live OpenAPI
- [Prepaid packages](https://provider.omnidev.tech/docs/packages): token packs, card checkout, API key use
- [Token bundles](https://provider.omnidev.tech/docs/bundles): sizes, prices, bonus tokens on agent rails, balance and usage endpoints, expiry
- [Status](https://provider.omnidev.tech/status): health snapshot and model verification state
- [Payments and nonces](https://provider.omnidev.tech/docs/payments): 402 challenge, settle-after-serve, X-Omni-Client-Nonce, idempotent replay
- [Full text for LLMs](https://provider.omnidev.tech/llms-full.txt): the pages above in one file

## Policies
- [Terms of service](https://provider.omnidev.tech/terms)
- [Privacy](https://provider.omnidev.tech/privacy)
- [Security policy](https://provider.omnidev.tech/security)
- [Security contact details](https://provider.omnidev.tech/.well-known/security.txt)
- [Contact form](https://provider.omnidev.tech/contact)

## Optional
- [Changelog](https://provider.omnidev.tech/changelog)