ModelBest

AGI for Lives

[ModelBest @modelbest_ai]

🚀 PilotDeck 2.0 is now open source.
✨ Create, check and revise Word, Excel & PPT files
✨ Long-running tasks. Intelligent model routing.
✨ Choose your models, deploy locally
An AI agent workbench that turns source materials into editable deliverables.
🔗 GitHub: https://t.co/mitJSSplKK
Web: https://t.co/KFEIHEJ31e

[ModelBest @modelbest_ai]

A closer look at MiniCPM-V 4.7 1B in action 🔎
@mfranz_on tests MiniCPM-V 4.7 1B on 11px contract text, table reconstruction, and a blurry, tilted label. See his results below 👇

[Marco Franzon @mfranz_on]

Small models are awesome!

The newest OpenBMB model
MiniCPM-V 4.7 1B read 11px contract text, rebuilt a table cell-perfect and nailed a blurry, tilted label. On reading, it keeps up with its 35B sibling.

Counting still trips it up, but wow. More tests are coming

[ModelBest @modelbest_ai]

MiniCPM-V 4.7 in a parallel coding test 🙌

@Tech2Wild reports running 16 coding agents concurrently with MiniCPM-V 4.7 35B-A3B in INT4, paired with a DSpark drafter. Explore the test results below 👇

[Tech2Wild @Tech2Wild]

16 Parallel Agents At The SAME TIME !

🚀 Minicpm-v-4.7 on MiniCPM-V-4.7-35B-A3B int4 + DSpark draft — 16 parallel coding agents
• Total: 59,459 tokens in 29s
• Aggregate: 2192.9 tok/s sustained · 2067.7 wall-avg · 2605.4 peak burst
• Per-stream decode: 142.1 high / 133.7 low / 137.1 avg tok/s
• TTFT 0.27s avg · E2E 27s avg

[ModelBest @modelbest_ai]

MiniCPM-V 4.7 in a coding workflow 💻
@KyleHessling1 shares a brief test in Claude Code, reporting that the model created a 3D scene and fixed its own errors. Take a look below 👇

[Kyle Hessling @KyleHessling1]

Mini CPM-V 4.7 brief test!

Numbers from the 5090 rig: Q5_K_M, 214 tok/s, 11.2K tok/s prompt processing, 262K context in 30.1 GB, 0.4 s vision.

It produced a very decent 3d scene in Claude Code given the speed and model size. It was also able to fix its own errors autonomously!

Most interestingly though, it can be run much faster offloaded than prior 35B MOE models. Looking forward to trying it on my MacBook and perhaps even my iPhone!

If you're wanting a new Qwen 35B, this might be a fun one to try in the interim!

[ModelBest @modelbest_ai]

Read the label. Do the math. Right on your phone.
In this demo, MiniCPM-V 4.7 1B runs locally on the iPhone, reads a nutrition label, and works out intake for the portion you ask about. It shows the calculation step by step.
Watch it in action 👇

[ModelBest @modelbest_ai]

Can you name the tallest building in this hazy skyline?👀
MiniCPM-V 4.7 35B-A3B runs smoothly on iPhone and iPad, identifying Beijing’s China Zun (CITIC Tower) and describing its distinctive shape.
Watch the demos 👇

[ModelBest @modelbest_ai]

How does MiniCPM-V 4.7 35B-A3B handle coding, SQL, vision, physics, and translation? 👀
@fahdmirza covers local setup on a single H100 and puts the model through five tests. Watch his walkthrough and verdict below 👇

[Fahd Mirza @fahdmirza]

💥 MiniCPM-V 4.7 35B-A3B just dropped, and we put it through five real tests on an H100

📱 Open-source MoE built for on-device use, now running on a local GPU

🔹 Setup and serving on a single H100
🔹 Coding: a chimney cake fire simulation
🔹 SQL: find the hidden bug and fix it
🔹 Vision: a handwritten note in three languages
🔹 Physics: reasoning over a page of equations
🔹 Translation: one Brazilian headline into 80 languages

🔥 Full test and honest verdict in the video below 👇

https://t.co/J2NeNhDSdC

@OpenBMB

[ModelBest @modelbest_ai]

Two MiniCPM-V 4.7 models, 13 benchmarks. 👀

@DogukanUrker tested 35B-A3B and 1B with BenchKit using our official API. Explore his results and takeaways on their strengths and limitations below 👇

[Doğukan @DogukanUrker]

MiniCPM-V4.7 is out. i got api early access from @OpenBMB and ran both sizes through 13 benchmarks with BenchKit.

the 35B-A3B is a really strong vision model. 86 on MMVP, 91 on OCRBench, 90.4 on ChartQA. it's also the best three.js scene writer i've tested so far (88 vs 80 for Empero’s qwen3.8-35b-a3b).

the 1B gets 82 on OCRBench and 85 on ChartQA. that's wild for 1.3B. it can't write full programs though.

coding and agentic tasks are its weak side. qwen3.6-35b-a3b is still ahead on HumanEval+ and PatchEval.

all scores first attempt, official API. weights config coming later this day.

[ModelBest @modelbest_ai]

A great hands-on look at MiniCPM-V 4.7 1B from @fahdmirza!

His walkthrough covers local setup with Ollama and llama.cpp, along with successes and failures from his tests. Check out the video below 🙌

[Fahd Mirza @fahdmirza]

🚀 Tiny model, big test: MiniCPM-V 4.7 1B, hands-on

📱 A one billion parameter vision model that fits on a phone

🔹 Installed the GGUF with llama.cpp and Ollama
🔹 Read an analog clock from a photo
🔹 Turned that clock photo into working HTML code
🔹 Solved a quadratic from a school math sheet
🔹 Handwriting translation across dozens of languages
🔹 Funny scene test: spotting body language and backstory

🔥 Honest results, wins and fails, in the full video below 👇

https://t.co/jGV90CrZKS

@OpenBMB

[ModelBest @modelbest_ai]

🚀 Announcing MiniCPM-V 4.7, a new open-source multimodal model series.
🔹 MiniCPM-V 4.7 35B-A3B brings SOTA performance to local devices through LES.
🔹 MiniCPM-V 4.7 1B — Lightweight, efficient multimodal AI. Easier to deploy.
Advancing the Performance–Efficiency Pareto Frontier.
Download, deploy, and build something we haven't imagined yet. 👀👇
🔗 GitHub: https://t.co/ckUnZG3rU1
🤗 Hugging Face: https://t.co/r0pM4rKO3o

[OpenBMB @OpenBMB]

🚀 Introducing MiniCPM-V 4.7, our new open-source MLLM series!
Advancing the Performance–Efficiency Pareto Frontier.
It brings strong multimodal capabilities to iOS and Android devices with extreme efficiency.
Two new models:
🔹 MiniCPM-V 4.7 35B-A3B — Bring SOTA performance to local devices with newly introduced LES (Lookahead Expert Scheduling) technique, a new milestone of on-device multimodal AI.
🔹 MiniCPM-V 4.7 1B — Lightweight, efficient multimodal AI. Easier to deploy.

⚡ What's new?
SOTA performance: Competitive results across 34 benchmarks spanning multimodal understanding, STEM reasoning, OCR and long-video comprehension.
Fast on-device inference: LES predicts and preloads expert weights to reduce inference stalls, boosting 5× speedup on iPad Pro M4 (16GB) with byte-for-byte identical outputs. It enables MiniCPM-V 4.7 35B-A3B to achieve 17+ tokens/s on iPad Pro M4 (16GB) and 13+ tokens/s on iPhone 17 Pro Max, while costing only 3GB memory 🤯
Efficient visual processing: 4× and 16× hybrid visual token compression. Even at 16× compression, 35B-A3B scores 90.4 on OCRBench, 5.2 points ahead of Qwen3.6 35B-A3B.
Easier customization: We introduce Self-Replay Fine-Tuning, a simple way to customize new target capabilities at low cost while maintaining general performance.
Broad hardware support: Day-0 adaptations for Apple, Intel, AMD, Arm and Rockchip, with many more platforms supported through FlagOS!

Download, deploy, and build something we haven't imagined yet. 👀👇
🔗 GitHub: https://t.co/msjm2QlrvX
🤗 Hugging Face: https://t.co/W95QKGhymw

[ModelBest @modelbest_ai]

How can audio–visual captions be reliably evaluated and errors pinpointed?

Explore OmniCapBench and its findings in @zhongyuyang021’s thread.

[Zhongyu Yang @zhongyuyang021]

🎬 Excited to share our NeurIPS 2026 research: OmniCapBench – a comprehensive benchmark for evaluating omnimodal video captioning models!

🌐 Homepage: https://t.co/zvzFzhbsai
📄 Paper: https://t.co/7Dz1pG1QlB
🤗 HF Papers: https://t.co/wn11umz7UX
⭐ Data: https://t.co/qthsxFUUA6

[ModelBest @modelbest_ai]

🚀Now live.
MiniCPM5-2B — 2B parameters, ranked first among open-source base models under 4B on Artificial Analysis.
A further validation of the densing law at the edge.

[OpenBMB @OpenBMB]

🚀 Meet MiniCPM5-2B, a 2B-parameter language model bringing high intelligence density to the edge, now open source!

It ranks #1 among open-source models under 4B parameters on the
@ArtificialAnlys Intelligence Index, with a score of 23.

It also scores 20 on the Agentic Index, bringing an early form of general-purpose agent capability to the edge.

Across 34 benchmarks, MiniCPM5-2B achieves an average score of 53.9, covering coding, math, long-context understanding, tool use, and agentic tasks.

And this release goes beyond the model itself.

We’re opening up the data, training recipes, and RL stack behind MiniCPM5-2B.

🤗 Hugging Face: https://t.co/FZOMTZhBjq
💻 GitHub: https://t.co/2L0I8bYv8f
Modelscope: https://t.co/WRlPNEAzgu
Web: https://t.co/gRSu65FeZL

[ModelBest @modelbest_ai]

Introducing the ModelBest "Ahead Four" Talent Program — for MS/PhD students graduating in 2027 or later.
• Equity grants before graduation
• Core work on MiniCPM
• 1:1 mentorship from senior scientists
• Serious compute
Apply: [email protected]
Toward AGI — Ahead Four.

[ModelBest @modelbest_ai]

✨ Thank you to @techreview for the kind mention! We're honored to be recognized among the emerging top Chinese AI startups to watch beyond DeepSeek.

https://t.co/jdlEfQuoQK

[ModelBest @modelbest_ai]

Excited by the amazing feedback at #CES2025! 🚀 Visit South Hall 3, Booth #41017, to explore MiniCPM 3.0’s on-device ChatGPT and MiniCPM-V 2.6’s GPT-4V. Let’s shape the future of edge AI together! ✨

[ModelBest @modelbest_ai]

🚀 Join us at #CES2025 in Las Vegas, Jan 7-10! 🌟
#ModelBest is showcasing our small-sized yet super-powerful edge model series – MiniCPM!
📍Meet us at South Hall, Booth #41017.

[ModelBest @modelbest_ai]

ModelBest Co. is honored to be on the Fortune 50 AI Innovators! Thanks to our partners and developers around the world—Let’s bring LLMs closer to users, and keep pushing the boundaries of AGI🚀🚀🚀
Appreciate @FortuneMagazine ！
https://t.co/aarOGFOCV5
#Fortune #minicpm #LLMs

[ModelBest @modelbest_ai]

further than MoE🤠

[OpenBMB @OpenBMB]

🥰A comprehensive overview of the differences between dense models, MoE models, and Configurable Foundation Models 👇
🔗paper: https://t.co/ynlWvNCsIw

[OpenBMB @OpenBMB]

🥰A comprehensive overview of the differences between dense models, MoE models, and Configurable Foundation Models 👇
🔗paper: https://t.co/ynlWvNCsIw

[ModelBest @modelbest_ai]

Introducing MiniCPM🚀 family members：
Feb.1 📝MiniCPM 1.0(2.4B), 🖼️MiniCPM-V 1.0(2.8B)
-Apr. 11 📝MiniCPM 1.2B, 🖼️MiniCPM-V 2.0(2.8B)
-May 20 🖼️MiniCPM-Llama3-V 2.5(8B)
-Jul. 5 📝MiniCPM-S（1.2B）
-Aug. 6 🖼️MiniCPM-V 2.6（8B）
-Sep. 5 📝MiniCPM 3.0(4B)

https://t.co/7yOxJmKpJU