tweetindex
EN

DeepSeek ✓

@deepseek_ai · joined 18 Oct 2023

Unravel the mystery of AGI with curiosity. Answer the essential question with long-termism.

1 135 325Followers
0Following
183Posts total
41.9MViews on collected posts

Latest posts

@deepseek_ai Congrats DeepSeek on another epic release! Hopefully you guys will release smaller models for people to run locally. 🙏🐋 It's great that DeepSeek-V4.1 has 196B engram making it more accessible. 77.1K views · 1K likes · 21 reposts · 21 replies Open on X →
🌐 Supporting open source. Expanding deployment options. We’ll work closely with the open-source community on V4.1-Flash inference support and explore more deployment options. Planning a large-scale deployment with 2,000 GPUs + a storage cluster? Let’s talk. 🔹 Model: 348.6K views · 2.9K likes · 212 reposts · 68 replies Open on X →
💰 More efficient architecture. Lower API prices. V4.1-Flash lets us serve more users at a lower cost. We’re passing the savings on to you. 🔹 Peak/off-peak pricing continues to balance demand. 🔹 Off-peak rates are 50% of peak rates. Schedule flexible workloads off-peak to https:
247.7K views · 1.9K likes · 53 reposts · 25 replies Open on X →
⚡ V4.1-Flash is now live on the DeepSeek API with native multimodal support. Set your model to deepseek-flash. 🔹 V4-Flash & V4-Flash-Vision-Exp are retired. For compatibility, deepseek-v4-flash and deepseek-v4-flash-vision-exp temporarily route to V4.1-Flash. 🔹 Tests by 410.8K views · 2K likes · 53 reposts · 23 replies Open on X →
💾 Smaller KV cache. Bigger savings. Compared with the previous generation, V4.1-Flash’s KV cache needs just: 🔹 1/4 the HBM 🔹 1/8 the SSD storage Cache-hit charges often account for a large share of agent costs. Compressing the cache cuts those costs significantly. 3/6 https://t
826.8K views · 3.1K likes · 193 reposts · 32 replies Open on X →
🧠 Asymmetric architecture. More intelligence, less cost. 🔹 552B-parameter MoE. 🔹 New Causal Encoder–Decoder architecture: just 8B active parameters for input, 16B for output. 🔹 New pre-training methods + larger-scale RL post-training deliver benchmark results ahead of https://t.
1.7M views · 4.9K likes · 354 reposts · 112 replies Open on X →
🚀 Introducing DeepSeek-V4.1-Flash: smarter, faster, more efficient. 🔹 Introducing the smallest model in our new architecture family, with native visual understanding. 🔹 Designed for greater capability, faster inference, higher throughput, and scaling to larger models. 1/6 https
6.8M views · 28.5K likes · 3.1K reposts · 1.1K replies Open on X →
DeepSeek-V4-Flash-Vision-Exp is now live on the DeepSeek API Platform! 🚀 🔹 This experimental multimodal model matches DeepSeek-V4-Flash on text capabilities—including agents, reasoning, and world knowledge. 🔹 On multimodal agent benchmarks, V4-Flash-Vision-Exp makes a major http
2.7M views · 11.5K likes · 1.2K reposts · 559 replies Open on X →
@deepseek_ai another one https://t.co/RmdyKyoqvo
17.1K views · 208 likes · 7 reposts · 0 replies Open on X →
@deepseek_ai You should launch a coding sub plan now. You got harness, you got model, and you have enough compute. Why not try a coding plan now. This will be a good entry in current tech dominant userbase!! 64.7K views · 350 likes · 4 reposts · 23 replies Open on X →
@deepseek_ai DeepSeek open-sourcing everything so we don’t have to live under the reign of Anthropic. https://t.co/Nk4fCEmh3J
GIF
90K views · 1.4K likes · 24 reposts · 16 replies Open on X →
🧩 DeepSeek Harness v0.1 is now available in Developer Preview! 🔹 We’re opening it up to developers building agent harnesses worldwide and open-sourcing the codebase in MIT license. 🔹 Powered by the Cordis meta-framework, DeepSeek Harness is an agent harness built around one 4.5M views · 20.1K likes · 2.5K reposts · 846 replies Open on X →
@deepseek_ai Gemini doesn't even get put into these comparison charts anymore it just sits there looking at you like this https://t.co/W2i5dLMQZC
28.8K views · 532 likes · 2 reposts · 10 replies Open on X →
API pricing update 💰 With the V4 lineup release, we’re updating our API pricing and introducing peak and off-peak rates. Off-peak rates are 50% lower than peak, enabling more flexible workload scheduling. 📉 New pricing takes effect at 16:00 UTC, Aug 16, 2026 🕒 https://t.co/l5CF
614.1K views · 1.8K likes · 191 reposts · 203 replies Open on X →
We’re launching DeepSeek-V4-Pro today! 🚀 🔷 Major Agent upgrades with strong production gains! 🔷 Flexible reasoning effort for V4-Pro & V4-Flash: low for simple tasks, high for daily Agent workflows, max for complex tasks. 🔷 Native OpenAI Responses API support, optimized for http
3M views · 15.3K likes · 1.8K reposts · 618 replies Open on X →
@deepseek_ai If DeepSeek-V4-Flash is this good and this small, imagine DeepSeek-V4-Pro! 🤯 And imagine running Flash locally on your own device! 🔥 We can't wait to make quants so everyone can run it locally!!! 159.3K views · 2K likes · 61 reposts · 57 replies Open on X →
⚠️ Note 🔷 DeepSeek-V4-Flash-0731 keeps the exact same model architecture and size as the preview version. 🔷 Today's upgrade applies ONLY to the DeepSeek-V4-Flash API. The DeepSeek-V4-Pro API and App/Web models remain unchanged for now. The official release of DeepSeek-V4-Pro 543.9K views · 3.8K likes · 139 reposts · 92 replies Open on X →
🚀 DeepSeek-V4-Flash Official API is now LIVE in public beta! 🔷 We’ve massively upgraded its Agent capabilities—benchmark scores are now far surpassing the V4-Pro-Preview. Check out the massive performance leap below! 👇 🔷 The official V4-Flash now natively supports the https://t.
9.6M views · 30.1K likes · 3.5K reposts · 1.7K replies Open on X →
@deepseek_ai Take all my data guys https://t.co/ExvdwPFPWE 46.1K views · 952 likes · 11 reposts · 8 replies Open on X →
@deepseek_ai China can have all my data thank you guys I welcome our new Chinese overlords with open arms 86.4K views · 2.1K likes · 39 reposts · 33 replies Open on X →
@deepseek_ai Capitalism: “AI for everyone!” Also capitalism: “That’ll be $200/month.” China: “Comrade, here’s your discounted AI.” Finally, socialism delivers something useful: affordable AI for the poor and needy while Silicon Valley charges rent for every token. 116.3K views · 1.7K likes · 47 reposts · 31 replies Open on X →
We are making our discount permanent! 🎉 Enjoy building with DeepSeek-V4-Pro and bring your innovative ideas to life! 🚀 https://t.co/V8atbTaogH
7.1M views · 24.4K likes · 2.7K reposts · 1.5K replies Open on X →
The DeepSeek-V4-Pro discount has been extended until May 31, 2026, 15:59 UTC! https://t.co/kd0taLhZhH
2.7M views · 6.8K likes · 557 reposts · 321 replies Open on X →

Against accounts of the same size

18 posts from the last 90 days, next to the 1M–10M follower range. shown to more people than peers of the same size.

Median views477 328this account14 648median for 1M–10M
Reach, %42.04%this account0.74%median for 1M–10M
Engagement, %0.59%this account0.70%median for 1M–10M
MetricThis accountMedian for 1M–10MRatio
Median views per post477 32814 64832.6×
Reach (views ÷ followers)42.04%0.74%56.8×
Engagement rate0.59%0.70%0.85×

Others in this range →   Compare with another account →   How these benchmarks are built →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

614.1K13 Aug
28.8K
4.5M
90K
64.7K
17.1K
2.7M21 Aug
6.8M10 Sep
1.7M
826.8K
410.8K
247.7K
348.6K
77.1K

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.41%13 Aug
1.89%
0.55%
1.64%
0.58%
1.26%
0.52%21 Aug
0.50%10 Sep
0.32%
0.42%
0.51%
0.80%
0.91%
1.42%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 1M–10M accounts is 0.70%.

What the audience does

Likes72.6%167 456 in total
Reposts7.3%16 842 in total
Replies3.2%7 377 in total
Quotes3.9%8 969 in total
Bookmarks13.0%29 890 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

Followers by day

19 Sep22 Sep

Daily snapshots since 19 Sep 2026; the dashed line is the starting count.

Similar accounts