Reflection launches Beam, Mistral debuts Large 4 as Western open models push back
Reflection, the US startup billed as an "American DeepSeek", released its first open-weight model Beam: a 501B-parameter MoE with 23B activated per token for coding, reasoning and agent workloads, whose final RL run used more than 10,500 Nvidia GB300 GPUs and over 100 million rollouts. A day later Mistral launched Mistral Large 4, its largest model yet at 1.05T parameters (52B active), 1M-token context and native multimodality, trained from scratch on about 3,800 Grace Blackwell GPUs in Europe; it is in API preview with full weights due in late October. Both companies concede Chinese open models remain ahead: Beam trails GLM-5.3, Kimi K3 and DeepSeek V4.1 Flash on DeepSWE v1.1 (44.4), Terminal Bench v2.1 (80.1) and HLE (36.2), so Reflection is pitching intelligence per unit of inference compute instead.
The US and Europe now have fresh frontier open-weight entries in the same week; still roughly a generation behind China's best, they pitch private deployment and inference efficiency, which could shape how enterprises pick long-term model providers.