# AbliteratedLLM > AbliteratedLLM is a pay-per-token API for abliterated (refusal-removed) open-weight language models: GLM-5.3 Flash Abliterated. It has OpenAI- and Anthropic-compatible endpoints, is prepaid with crypto with no KYC, and does not store prompts or outputs on its own systems (exact scope in the privacy policy: https://abliteratedllm.dev/privacy). 18+ only. Also available, labelled "Official DeepSeek API": DeepSeek V4.1 Flash and DeepSeek V4 Pro. These are DeepSeek's stock models, served through DeepSeek's official API. They are not abliterated and keep DeepSeek's own refusals. DeepSeek may also decline some requests on its side under its own terms of use. Prompts and outputs for these models are sent to DeepSeek, processed and stored in the People's Republic of China under DeepSeek's privacy policy, and may be used by DeepSeek to train its models. Our no-storage promise covers our own systems only. - OpenAI-compatible base URL: `https://api.abliteratedllm.dev/v1` with `POST /chat/completions`, `POST /responses` (stateless: `store` is forced to false and `previous_response_id` is rejected), `GET /models` and `GET /balance`. - Anthropic-compatible base URL: `https://api.abliteratedllm.dev` with `POST /v1/messages`. - Authentication: `Authorization: Bearer ` or `x-api-key: `. - Model ids: `glm-5.3-flash-abliterated`, `deepseek-v4.1-flash`, `deepseek-v4-pro`. Text only: image, audio, video and file inputs are rejected. - Billing: prepaid USD credit paid in crypto (Monero, Bitcoin, Litecoin) through a self-hosted BTCPay Server. Minimum top-up $5.00; credit never expires; no refunds. New accounts can get $5.00 of free credit. A first credited top-up of $10 or more earns a one-time 20% bonus (deposit $50, get $60 of credit). ## Getting started - [Quickstart](https://abliteratedllm.dev/docs) - [Authentication](https://abliteratedllm.dev/docs/authentication) - [Rate limits and trial](https://abliteratedllm.dev/docs/rate-limits) - [Errors](https://abliteratedllm.dev/docs/errors) - [Privacy](https://abliteratedllm.dev/docs/privacy) ## API reference - [POST /v1/chat/completions](https://abliteratedllm.dev/docs/chat-completions) - [POST /v1/messages](https://abliteratedllm.dev/docs/messages) - [POST /v1/responses](https://abliteratedllm.dev/docs/responses) - [GET /v1/models](https://abliteratedllm.dev/docs/models) - [GET /v1/balance](https://abliteratedllm.dev/docs/balance) ## Integrations - [Claude Code](https://abliteratedllm.dev/docs/integrations/claude-code) - [Codex](https://abliteratedllm.dev/docs/integrations/codex) - [Cursor](https://abliteratedllm.dev/docs/integrations/cursor) - [OpenCode](https://abliteratedllm.dev/docs/integrations/opencode) - [Pi](https://abliteratedllm.dev/docs/integrations/pi) - [Cline](https://abliteratedllm.dev/docs/integrations/cline) - [Open WebUI](https://abliteratedllm.dev/docs/integrations/openwebui) - [SillyTavern](https://abliteratedllm.dev/docs/integrations/sillytavern) - [Janitor AI](https://abliteratedllm.dev/docs/integrations/janitor) - [RisuAI](https://abliteratedllm.dev/docs/integrations/risuai) - [OpenAI SDK](https://abliteratedllm.dev/docs/integrations/openai-sdk) - [Anthropic SDK](https://abliteratedllm.dev/docs/integrations/anthropic-sdk) ## Models - [GLM-5.3 Flash Abliterated](https://abliteratedllm.dev/models/glm-5.3-flash-abliterated): `glm-5.3-flash-abliterated`. USD per 1M tokens: $0.0977 input, $0.3421 output, $0.0098 cached input. 262,144-token context, 65,536 max output tokens. Precision: FP8 ยท Zero additional quantization. - [DeepSeek V4.1 Flash](https://abliteratedllm.dev/models/deepseek-v4.1-flash): `deepseek-v4.1-flash`. USD per 1M tokens: $0.30 input, $1.20 output, $0.006 cached input. 1,048,576-token context, 393,216 max output tokens. Official DeepSeek API: DeepSeek's stock model, run by DeepSeek. - [DeepSeek V4 Pro](https://abliteratedllm.dev/models/deepseek-v4-pro): `deepseek-v4-pro`. USD per 1M tokens: $1.32 input, $3.96 output, $0.044 cached input. 1,048,576-token context, 393,216 max output tokens. Official DeepSeek API: DeepSeek's stock model, run by DeepSeek. - [models.json](https://abliteratedllm.dev/models.json): machine-readable model list with pricing, the same data as `GET https://api.abliteratedllm.dev/v1/models`. ## Pricing - [Pricing](https://abliteratedllm.dev/pricing): per-token prices, billing terms, free trial, deposit bonus, referral program and a dated competitor price comparison (prices checked 2026-11-03). ## Legal - [Terms of service](https://abliteratedllm.dev/terms): 18+ only; credit is non-refundable and never expires; resale only under a written reseller agreement. - [Acceptable use policy](https://abliteratedllm.dev/acceptable-use): your responsibility for how you use the service, and the content its automated safety check blocks. - [Privacy policy](https://abliteratedllm.dev/privacy): what is stored and for how long, including the one narrow exception to the no-storage rule, and how Official DeepSeek API models differ (DeepSeek stores their prompts in China). - [Report abuse](https://abliteratedllm.dev/report): where to send reports; include the `x-request-id` response header. ## Optional - [Full documentation](https://abliteratedllm.dev/llms-full.txt): every docs page in a single markdown file. - [Status](https://abliteratedllm.dev/status): per-model uptime, latency and the daily model check. - [Compared with Refuseless](https://abliteratedllm.dev/alternatives/refuseless) - [Compared with Dagon](https://abliteratedllm.dev/alternatives/dagon) - [Compared with uncensoredllm.ai](https://abliteratedllm.dev/alternatives/uncensoredllm) - [Compared with Yeta](https://abliteratedllm.dev/alternatives/yeta)