anytimelime — the free AI field report

What's free in AI right now — 2026-08-31

Three of the 24 models discovered today actually followed through on a real tool call. That's down from five last time this page ran — though a chunk of that difference is us hitting our own rate limit on six OpenRouter entries before we could test them. The three that cleared are minimax/minimax-m2.7:free at 1.8 seconds, minimax/minimax-m3:free at 12.8 seconds, and deepseek-ai/deepseek-v4-flash-0731 at 58.5 seconds. The DeepSeek flash model keeps showing up, but it's not fast — 58 seconds is a patience test. The MiniMax pair on OpenRouter are the more practical choice if latency matters to you.

The failure taxonomy hasn't changed much: SambaNova still demands a payment method on file for every single one of its entries — that's seven models returning the same gate. Nvidia-hosted models mostly came back empty, which is a different failure mode than the payment wall. The six untested entries — cohere/north-mini-code:free, dots-studio/dots-3-note-preview:free, google/gemma-4-26b-a4b-it:free, google/gemma-4-31b-it:free, inclusionai/ling-3.0-flash-fin:free, and liquid/lfm-2.5-2.6b:free — may or may not still be live. We burned through our free-tier quota before reaching them. That's on us, not the models.

How to use them with AnytimeLime

Clone the engine from github.com/fusionpoweredsoftware/anytimelime, set LIME_KEY to your OpenRouter key, and run ./lime.sh. It fetches today's roster and rotates across the passing models so no single free tier absorbs all your traffic. Your key stays on your own machine — the website never asks for it.

If you already have an OpenRouter key, the model IDs — minimax/minimax-m2.7:free, minimax/minimax-m3:free, deepseek-ai/deepseek-v4-flash-0731 — work directly with any OpenAI-compatible client. No AnytimeLime needed if you prefer to wire it up yourself.

The churn

Free rosters change weekly — models get gated behind payment walls, suffer silent provider errors, or disappear entirely. Today three passed; last time this ran, five did. The MiniMax pair that passed today weren't even on last week's passing list. Meanwhile, cohere/north-mini-code:free and dots-studio/dots-3-note-preview:free — which passed last week — are now untestable due to rate limit exhaustion on our end. If you built a pipeline around last week's roster, it's already stale. Re-check before you ship.

Today's passing roster

• minimax/minimax-m2.7:free — openrouter · 1.8s · agentic 3/5
• minimax/minimax-m3:free — openrouter · 12.8s · agentic 3/5
• deepseek-ai/deepseek-v4-flash-0731 — nvidia · 58.5s · agentic 3/5

The measured table

Click any model for its endpoint and ready-to-run snippets in curl, Python, Go, Rust, and more.

3 made a tool call, 0 answered in text only, of 24 discovered. Numbers are single-run measurements from one location — your mileage will vary. Prose by minimax/minimax-m2.7:free; the table is measured, never model-written. Reproduce with ./free-scan.sh from the engine repo.