OpenRouter has added a new free model called inclusionAI: Ling 3.2 Flash. The model runs on a single provider but OpenRouter will automatically switch to another if the first fails. It delivers 32 tokens per second and a round‑trip latency of 1.94 seconds, which is the best among the providers listed. Availability over the past three days is recorded at 98.82 %. The API works with any OpenAI‑compatible SDK by changing the base URL and model slug.
Why it matters
Developers can try a high‑throughput, low‑latency LLM for free without worrying about downtime, speeding up prototyping of chat or text‑generation apps.
