Reflection AI dropped the open weights of its first model today.
AspireAI community · huggingface.co
Reflection AI dropped the open weights of its first model today.
Beam 1.0 is a 501-billion-parameter mixture-of-experts model with 23 billion active per token, a 1-million-token context window, and Apache 2.0 licensing. Weights are live on Hugging Face in BF16, FP8, and NVFP4 formats with a speculative-decoding drafter.
The big news is the ecosystem. Day-zero support covers Hugging Face Transformers, vLLM, SGLang, NVIDIA Dynamo, Unsloth, and Axolotl, plus 40-plus partners including AWS, Azure, Google Cloud, Cloudflare, Fireworks, Together, OpenRouter, Ollama, OpenCode, LM Studio, and Hugging Face. On-prem ships through Dell and HPE. Reflection also open-sourced Mirror CLI, the Rust harness it trained Beam on.
Think of it as: download a frontier-scale model, run it on whatever stack you already use, host it on whatever cloud you already trust.
Honest caveat: the comparable-to-GLM-5.2 benchmark claim comes from Reflection itself, not yet replicated.
Read the story and explore more from the AspireAI community.
Two critical PHP remote code execution flaws — one found by GPT-5.6 Sol — were patched on October 21, 2026, and OpenAI published a disclosure framework alongside the fix. Fixed versions: PHP 8.5.12, 8.4.27, 8.3.36 and 8.2.35. Both allow unauthenticated remote code execution, and PHP's advisory says risk is highest for mod_php, where a demonstrated exploit triggers before any application code runs. The CVEs are distinct failure modes. CVE-2026-109618 (OOM corruption): out-of-memory handling can interrupt PHP while internal request data is only partially initialized, and request cleanup may then treat the invalid data as valid. CVE-2026-109619 (stale global state): request-owned internal state can survive after its backing memory is freed, and a later request served by the same worker may use that stale state. OpenAI states GPT-5.6 Sol identified the first flaw through its Patch the Planet initiative, and that Wiz independently discovered and reported both vulnerabilities. PHP's maintainers led the fixes and release, with Wiz, Cloudflare and OpenAI contributing research and coordination. Wiz's write-up adds that it used GPT-5.6 Sol and Gemini 4 Argon throughout, with Gemini 4 Argon contributing to finding the vulnerability and validating the patch. OpenAI describes the gap between discovery and public exploitation as a "Defender's Window," proposes a five-stage response framework, and committed $1 billion in subsidised Daybreak model access for organizations with limited security resources. Honest caveats: these are OpenAI's and Wiz's own characterizations, and OpenAI also notes the same capabilities help attackers build offensive tooling faster. PHP 8.1 and older no longer receive updates from the PHP project.