the first open-weights model from Reflection, a 501B-total, 23B-active mixture-of-experts built for agentic tool use, coding and general reasoning. it reads text only and its weights ship under Apache 2.0.
pick it if
you want a long-window text model priced with the flash tier
you want Apache 2.0 weights, with BF16, FP8 and NVFP4 builds to self-host
you run tool-calling agents on text and repeat long prompts, so cached input matters
skip it if
you need image, audio or video input; Beam reads text only
long autonomous coding is the job; its reported DeepSWE v1.1 score sits below GLM-5.3 Flash and DeepSeek V4.1 Flash
Inception's Mercury 3 is a text-only diffusion llm with selectable reasoning effort and a 1M-token context window. it replaces Mercury 2.5 at lower input and output rates, with the window up from 260K tokens.
pick it if
you need low-cost text work across contexts up to 1M tokens
latency matters; Inception cites 1,068 tokens/s on NVIDIA GPUs
your prompts repeat, so cached input bills at a tenth of input
skip it if
you need image input
you need stable long-term rates; the launch discount has no announced end date
faq
which is cheaper, Beam or Mercury 3?
Mercury 3 lists lower input and output prices. on a coding-agent session (2M tokens in, 60,000 out) Beam costs $0.27 and Mercury 3 $0.22.
can i switch between Beam and Mercury 3 mid-task?
yes. in an octomind session, /model octohub:beam and /model octohub:mercury-3 switch models without losing the conversation, so you can draft on one and review on the other.
try either model
switch between models inside the same octomind session.