A new open-weights DeepSeek V4.1-Flash competitor just appeared.
It fits on 2x DGX-Sparks and can ...
operate scientific software, check its own work, and FIX its mistakes.
It uses the same architecture as GLM-5.3-Flash.
This is Sage-S2.1-Flash from Shanghai AI Lab.
Here are the model's stats ...
🧠 320B total / 18B active
📚 ~1M token context
👁️ Multimodal
🔓 MIT license
💾 BF16 + FP8 weights downloadable
Shanghai AI Lab demonstrated it working with 12,487 mixed medical images.
Example application, the model
👉 Identified 139 relevant CT slices
👉 Attempted a reconstruction in 3D Slicer
👉 Detected a slice-ordering inconsistency
👉 Corrected the error
👉 Reconstructed the skeleton in 3D
It didn't just describe the images.
Also, it operated scientific software and used verification feedback to correct its work.
And Sage supports 10+ types of scientific inputs
🧬 DNA / RNA / protein sequences
🧪 Molecular graphs
💎 Crystal structures
📈 Spectra
🧠 EEG / ECG signals
Now look at these developer-reported benchies.
MLE-bench Lite:
🔥 Sage S2.1 Flash → 90.3
🐋 DeepSeek V4.1 Flash → 89.2
🟠 Claude Opus 5 → 88.0
⚫ GLM-5.3-Flash → 84.3
🟣 Qwen3.8-Max → 82.2
🟢 GPT-5.6 Sol → 80.7
And compared with GLM-5.3-Flash:
OfficeQA Pro
64.7 → 72.2
Harvey LAB
20.0 → 25.0
SpreadsheetBench 2
32.4 → 37.1
Terminal-Bench 2.1
84.3 → 87.9
Not a clean sweep across every benchmark, but these are interesting results for a scientific agent.
⚖️The weights are already here.
💾 BF16 → ~645GB
💾 Official FP8 → ~330GB
Raw weight math:
4-bit → ~160GB
3-bit → ~120GB
2-bit → ~80GB
Remember all the work Localmaxxers have done getting GLM-5.3-Flash running on affordable hardware?
So who's going to quantize it for a 192GB Gorgon Halo?
Or two DGX Sparks? 👀
Imagine running your own scientific research agent that doesn't just generate answers.
It executes experiments, checks results, and corrects its mistakes.
Link in reply
@dorszafranski Ale to już od dawna nie opłaca się centrować divów najlepszymi modelami. :) W Claude polecam combo Opus z Haiku, który jest niedocenianym hidden gemem.
👾HD-2D style house built with Gemini Hydrogen High.
I really liked how Hydrogen came up with the idea of adding its own testing panel and used it to click through and test different effects during development.
Something's a bit off with the lighting, but overall the result is pretty cool!
On Google AI Pro plan it consumed 4% of 5h limit and 2% weekly which is really good.
Done in 8 minutes.
Codex still have many advantages over opus. It can be used safely as brain for Hermes, it do almost everything what I ask for like deleting users from db. Opus refuses to do many things.
Can’t wait for sol 6.2 or astra. If they will be 10-15% weaker than Opus I will be so happy.
@dorszafranski Nauka architektury to też nie jest najlepsza inwestycja chyba że ktoś celuje w topowe największe projekty. Te małe i średnie ai już dobrze projektuje a będzie tylko lepiej. Kluczem jest umiejętność zaawansowanej pracy z ai, dobór modeli, harnessow itd