Ashish Jain

Ashish Jain | Founder, Probe India | AI Learner | Cloud & Network Architect | Cyber Security & Automation Enthusiast

[Joseph Miclaus @josephmiclaus]

GPT-5.6 Luna Max is actually OP.

To be able to use Max you've got to enable it in your Codex setting 👇

[谢小总 @xie_Dl]

Codex极致省钱大法

原来 Codex 可以开 Luna Max 子代理

GPT-5.6 sol负责决策

GPT-5.6 Luna负责搬砖

ChatGPT Plus套餐用出Pro 20x的效果

去设置-配置-可用推理强度把Max打开

接着把下面这段提示词发给 Sol：
-----
在以下路径创建一个名为 luna_worker 的自定义 Agent： ~/.codex/agents/luna-worker.toml 使用以下配置： model = "gpt-5.6-luna" model_reasoning_effort = "max" 请为这个 Agent 补充清晰的 description 和 instructions。 luna_worker 只负责处理范围明确、边界清晰、可以独立完成的委派任务，不负责修改整体任务目标，也不要自行扩大工作范围。 请保留我现有的其他 Codex 配置，不要覆盖或删除无关内容。 创建完成后： 1. 根据我当前安装的 Codex 版本检查配置格式是否兼容； 2. 展示本次修改产生的 diff； 3. 确认配置有效； 4. 后续需要调用子代理时，优先使用 luna_worker。
-----

这样Luna Max 就可以作为子代理搬砖省钱了。

[AshutoshShrivastava @ai_for_success]

GPT-5.6 Sol > Opus 5
End of discussion.

[Telegraph @Telegraphprotoc]

Calling all AI Engineers: the Telegraph Hackathon is open.

[IndiaAI @OfficialINDIAai]

What happens when AI meets the real challenges of governance?

Action. Innovation. Impact.

Following the India AI Impact Summit 2026, 762 problem statements were identified across 62 Ministries and Departments, covering opportunities in governance, public service delivery, operational efficiency and sector-specific challenges.
These are not simply numbers. They represent 762 opportunities to explore how AI can make public services smarter, more efficient and more responsive. IndiaAI is helping create pathways for AI-enabled interventions that can move from ideas and prototypes to pilots, deployment and scale.

AI isn’t just changing technology! It is changing what technology can do for public good.

🔗 Know more: https://t.co/yheM6yvgAA

#IndiaAI #AIForGovernance #AIForPublicGood #GovTech #DigitalIndia #ArtificialIntelligence #AIInnovation #PublicServiceDelivery #TechForGood #AIForIndia

@AshwiniVaishnaw @JitinPrasada @SecretaryMEITY @sudeep_tweets @_DigitalIndia @mygovindia @PIB_India

[Tech Buzz China @TechBuzzChina]

Moonshot AI oversubscribes to $3.5B, targets $50B pre-IPO

Moonshot AI raised $3.5 billion in a round that far exceeded its initial target of $1–2 billion, per Bloomberg, valuing the Kimi developer at $35 billion post-money. The Beijing-based company is already talking to investors about another raise at a $50 billion pre-money valuation, eyeing a Hong Kong IPO as soon as this year.

Our read: The $3.5 billion round overshot its target directly because Kimi K3 generated buzz in Silicon Valley, per the source. The quick follow-on at a $50 billion pre-money valuation shows investors now see Moonshot as a frontier competitor. The company is moving fast to cash in on the model’s momentum. Now Moonshot has to make revenue and unit economics catch up to its valuation.

[NVIDIA AI Infrastructure @NVIDIAAIInfra]

Same hardware doesn't mean same performance.

NVIDIA Exemplar Cloud unlocked 8–12% training throughput gaps on two clusters running identical NVIDIA H100, GB200 NVL72, or GB300 NVL72 systems — not from the chips, but from how the stack is configured.

Our engineers traced four partner deployments to the root cause — and documented exactly what fixed each one.

🔗 Read the tech blog: https://t.co/Xl0tprX1rP

[Arena.ai @arena]

Inkling-Small (@thinkymachines) debuts at rank ~#88 (1431 pts, AutoEval) in Text Arena and ~#21 among open.

This makes it one of only five US models in the top 25 open models overall.

Note: this is an early AutoEval score, in which a Reward Model trained on Arena's human preference data casts automatic votes in place of live voting, giving a day-1 signal right at release. Scores are tagged preliminary until correlated with live votes. See the thread for more on how AutoEval works and the methodology behind it.

Congrats to @thinkymachines for another strong contribution to the open ecosystem!

We’ll continue to track to see where it lands in rank as live votes come in.

[Sam Altman @sama]

major price cuts today:

*80% drop for GPT-5.6 Luna, now $0.20 per million input tokens and $1.20 per million output
*20% drop for GPT-5.6 Terra, to $2/$12
*GPT-5.6 Sol gets Fast mode in the API, up to 2.5x the speed for 2x the price, same intelligence

[DeepSeek @deepseek_ai]

🚀 Day 2 of #OpenSourceWeek: DeepEP

Excited to introduce DeepEP - the first open-source EP communication library for MoE model training and inference.

✅ Efficient and optimized all-to-all communication
✅ Both intranode and internode support with NVLink and RDMA
✅ High-throughput kernels for training and inference prefilling
✅ Low-latency kernels for inference decoding
✅ Native FP8 dispatch support
✅ Flexible GPU resource control for computation-communication overlapping

🔗 GitHub: https://t.co/Q6eRxZ9kgW

[DeepSeek @deepseek_ai]

🚀 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 Responses API format and is fully adapted for Codex!

Check out the configuration details in our official API docs: https://t.co/smCwQZMeiq

[DeepSeek @deepseek_ai]

🔥DeepSeek Input Cache Price Drop!

Effective immediately, the price for input cache hits across the ENTIRE DeepSeek API series is reduced to just 1/10th of the original price! Build more efficiently for less.

📌Reminder: The DeepSeek-V4-Pro 75% OFF promotion is still active until May 5th, 2026, 15:59 (UTC Time).

[CNBC @CNBC]

China memory chipmaker CXMT skyrockets 470% in Shanghai debut https://t.co/D5xhlp16Fp

[iglooo101 @iglooo101]

@business https://t.co/ycTfAHO3S9

[Bloomberg @business]

BREAKING: CXMT shares surge 472% in a historic Shanghai debut after raising as much as $9.8 billion in China’s second-largest IPO https://t.co/AT5ACwchCa

[NCN Magazine @ncnmagazine]

Samsung Solve for Tomorrow Selects Top 100 Young Innovators with over 50% Representing Tier 2 & Tier 3 Cities

𝐊𝐧𝐨𝐰 𝐌𝐨𝐫𝐞👇
https://t.co/sYHf5a69O9

@SamsungIndia #Samsung @ncnmagazine #ncnmagazine #ncnonline #NCN

[NCN Magazine @ncnmagazine]

Vertiv Hosts Technovation 2026, Its Fifth Engineering Innovation Forum to Celebrate Excellence in Engineering at Vertiv

𝐊𝐧𝐨𝐰 𝐌𝐨𝐫𝐞👇
https://t.co/KiYDCpW5PU

@Vertiv #Vertiv @ncnmagazine #ncnmagazine #ncnonline #NCN

[Ashish Jain @ashishjain13]

Financial advise! cfur

[Ashish Jain @ashishjain13]

Invite friends!

[Ashish Jain @ashishjain13]

Test new things! yebf

[Ashish Jain @ashishjain13]

@elliotrades Airdrop Don't be late 🤯 https://t.co/ZcGQ1YZwEV.

[Ashish Jain @ashishjain13]

Invite friends! aerk

[Ashish Jain @ashishjain13]

Well!