I'm going to have to heavily disagree. Gemini 2.5 Pro has super impressive performance on large context problems. I routinely drive it up to 4-500k tokens in my coding agent. It's the only model where that much context produces even remotely useful results.
I think it also crushes most of the benchmarks for long context performance. I believe on MRCR (multi round coreference resolution) it beats pretty much any other model's performance at 128k at 1M tokens (o3 may have changed this).
I find that it consistently breaks around that exact range you specified. In the sense that reliability falls off a cliff, even though I've used it successfully close to the 1M token limit.
At 500k+ I will define a task and it will suddenly panic and go back to a previous task that we just fully completed.
Yet somehow chatting with Gemini in the web interface, it forgets everything after 3 messages, while GPT (almost) always feels natural in long back-and-forths. It’s been like this for at least a year.
I think it also crushes most of the benchmarks for long context performance. I believe on MRCR (multi round coreference resolution) it beats pretty much any other model's performance at 128k at 1M tokens (o3 may have changed this).