Endlesstokens

Endlesstokens is an OpenAI- and Anthropic-compatible API that provides unlimited AI tokens on a flat time pass.

Endlesstokens offers an OpenAI- and Anthropic-compatible API designed for coding agents, editors, and scripts without per-token billing. Users purchase a one-day or three-day time pass to run sessions with hosted open-weight models like Qwen3.8-27B or use et:auto to select models. The service supports up to 128K context windows, vision with up to five images per request, and tool calling. Each account includes two parallel sessions per key, and the company hosts its own GPUs rather than reselling other providers. It suits developers running heavy agent loops who want predictable costs without surprise bills or request limits.

Endlesstokens at a glance

Best for: Software engineer
Runs on: Web, API
Pricing: 1-day pass $15; 3-day pass $12 per day; top-ups available for active passes at 3 hours for $4.49 and 6 hours for $7.99
Released: October 2026
Category: AI coding tool
Website: endlesstokens.ai

Key features

Unlimited AI tokens for the duration of a pass
OpenAI- and Anthropic-compatible API
Support for up to 128K context windows
Image support up to 5 images per request
Two parallel sessions per key
Hosted open-weight models like Qwen3.8-27B
et:auto model selection for conversations

Use it for

Running automated coding agent loops without token meters
Executing terminal pair programming with Aider or Claude Code
Integrating unlimited AI inference into custom editor extensions
Handling tasks with vision and tool calling via API

Alternatives to Endlesstokens

OmniRoute — Routes requests across 352 providers with automatic fallback and cost-based selection.
OpenCode — Works with 75+ model providers, local models, and existing subscriptions
OpenTokenMini — Measures every agent call cost and links it to code commits.
AI/ML API — Access over 1000 AI models using one standard OpenAI-compatible API schema.
OpenRouter — Access over 500 models and 80 providers through one unified API.
FreeToken — Runs 290B+ MoE models on consumer gaming PCs using elastic edge resources.
OpenSandbox — Provides isolated environments with multi-language SDKs and lifecycle management.
Codex Infinity — Autonomous agents plan, test, and fix CI in isolated sandboxed environments.
TokenTracker — Monitor AI spending across 43 coding tools in one local, private dashboard.
7API — Access multiple top models and uncensored variants with pay-per-token pricing.
Wiyomi — Saves up to 70 percent on tokens while handling coding and research tasks.
Billion Context — Compresses conversation history to run single AI coding sessions for days.

All Endlesstokens alternatives