Beam 1.0 is Reflection's open-weight reasoning model for coding and long-horizon agentic workflows.
Prices update automatically — checked daily against provider list prices.
Independent benchmark scores for Beam 1.0 501B A23B, measured by Artificial Analysis. Higher is better.
Beam 1.0 501B A23B is an open-weight AI model by Reflection. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Beam 1.0 501B A23B is an AI language model from Reflection. It is open-weight: you can download it and run it on your own hardware, for free. It scores 17.7 on the Artificial Analysis intelligence index.
The weights are free and open — you can self-host Beam 1.0 501B A23B and pay nothing per token. If you prefer a hosted API, list prices are $0.25 per million input tokens and $1 per million output tokens.
Independent benchmarks from Artificial Analysis give it Humanity's Last Exam 34.7%, Long Context Reasoning 78%, SciCode 48.5%.
It generates about 396.6 tokens per second, with a median 0.48s delay before the first token. Measured independently by Artificial Analysis.
Yes. Beam 1.0 501B A23B has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.