Nitin Kaithvar

Founder | AI News & Updates | 550K+ Views on YouTube

[Nitin Kaithvar @Nitinkaithvar]

Dear 𝕏 algorithm,
Show this to the people who are curious about AI on 𝕏.

[Nitin Kaithvar @Nitinkaithvar]

@Zephyr_hg NEWS

[Nitin Kaithvar @Nitinkaithvar]

@jordan_ross_8F Marketing

[Nitin Kaithvar @Nitinkaithvar]

@PrakashS720 Banana

[Nitin Kaithvar @Nitinkaithvar]

@matt_elevenlabs 11b

[Nitin Kaithvar @Nitinkaithvar]

Just a reminder that none of this existed last year:

DeepSeek: R1
Qwen2.5-Max
Sonar Reasoning
Mistral: Mistral Small 3
o3 Mini
Mistral Small 3
Gemini 2.0 Flash-Lite
Gemini 2.0 Pro
Google: Gemini 2.0 Flash
Llama-3.1-70B-Instruct
OpenAI: o3 Mini High
Grok-3
Saba
Claude 3.7 Sonnet (thinking)
Claude 3.7 Sonnet
Gemini 2.0 Flash Lite
GPT-4.5
DeepSeek-V2.5
llama-4
Sonar Reasoning Pro
Sonar Pro
Sonar Deep Research
Phi 4 Multimodal Instruct
Qwen2.5-72B-Instruct
GPT-4o-mini Search Preview
GPT-4o Search Preview
Gemma 3 27B
Gemma 3 4B
Gemma 3 12B
Mistral Small 3.1 24B
Mixtral-8x22B-v1
o1-pro (api)
DeepSeek V3 0324
Gemini 2.5 Pro
DeepSeek-V3-0324
Llama-4-Scout
Llama-4-Maverick
Llama 3.1 Nemotron Ultra 253B v1
Grok 3 Mini Beta
Grok 3 Beta
GPT-4.1
GPT-4.1 Mini
GPT-4.1 Nano
o4-mini
o4 Mini High
o3
o4 Mini
Qwen3
Qwen3-32B
DeepSeek Prover V2
Phi 4 Reasoning Plus
DeepSeek-R1-Distill-70B
Mistral Medium 3
Google: Gemini 2.5 Pro Preview 05-06
Phi-4
Codex Mini
Gemma 3n 4B
Devstral Small 2505
Claude Opus 4
Claude Sonnet 4
DeepSeek-R1-0528
DeepSeek R1 0528 Qwen3 8B
Smaug-72B
Gemini 2.5 Pro Preview 06-05
o3 Pro
Grok 3 Mini
Kimi Dev 72B
MiniMax M1
Gemini 2.5 Flash
Gemini 2.5 Pro
Gemma-3-27B
Mistral Small 3.2 24B
Inception: Mercury
Mistral Large 2
Morph V3 Large
Morph V3 Fast
Grok 4
Gemma 3n 2B
Devstral Medium
Devstral Small 1.1
Kimi K2 0711
Google: Gemini 2.5 Flash Lite
GLM 4 32B
GLM 4.5
GLM 4.5 Air (free)
GLM 4.5 Air
Codestral 2508
Claude Opus 4.1
gpt-oss-120b
gpt-oss-20b
Claude Opus 4.1
GPT-5 Chat
GPT-5
GPT-5 Mini
GPT-5 Nano
GLM 4.5V
Mistral Medium 3.1
GPT-4o Audio
DeepSeek V3.1
DeepSeek V3.1
Grok Code Fast 1
Gemini 2.5 Flash Image Preview (Nano Banana)
Kimi K2 0905
NVIDIA: Nemotron Nano 9B V2
Qwen3-110B
Grok 4 Fast
DeepSeek V3.1 Terminus
GPT-5 Codex
Gemini 2.5 Flash Preview 09-2025
Gemini 2.5 Flash Lite Preview 09-2025
Claude Sonnet 4.5
DeepSeek V3.2 Exp
GLM 4.6
GLM 4.6 (exacto)
GPT-5 Pro
Gemini 2.5 Flash Image (Nano Banana)
o3 Deep Research
o4 Mini Deep Research
Llama 3.3 Nemotron Super 49B V1.5
GPT-5 Image
Claude Haiku 4.5
Claude Haiku 4.5
GPT-5 Image Mini
MiniMax M2
Nemotron Nano 12B 2 VL (free)
Nemotron Nano 12B 2 VL
gpt-oss-safeguard-20b
Sonar Pro Search
Voxtral Small 24B 2507
Nova Premier 1.0
Kimi K2 Thinking
KAT-Coder-Pro V1 (free)
GPT-5.1
GPT-5.1 Chat
GPT-5.1-Codex
GPT-5.1-Codex-Mini
Gemini 3
Google: Gemini 3 Pro Preview
GPT-5.1-Codex-Max
Grok 4.1 Fast
Nano Banana Pro (Gemini 3 Pro Image Preview)
Claude Opus 4.5
DeepSeek-V3.2
Mistral Large 3 2512
DeepSeek V3.2 Speciale
DeepSeek V3.2
Nova 2 Lite (free)
Nova 2 Lite
Ministral 3 14B 2512
Ministral 3 8B 2512
Ministral 3 3B 2512
GPT-5.1-Codex-Max
Relace Search
GLM 4.6V
Devstral 2 2512 (free)
Devstral 2 2512
GPT-5.2 Chat
GPT-5.2 Pro
GPT-5.2
MiMo-V2-Flash (free)
Nemotron 3 Nano 30B A3B (free)
Mistral Small Creative
Gemini 3 lite

[Eric Topol @EricTopol]

Generative AI, 3 years in

[Nitin Kaithvar @Nitinkaithvar]

But why shaped like UFOs

[NASA Ames @NASAAmes]

NASA’s DiskSat is gearing up to launch into low Earth orbit!

Four plate-shaped spacecraft will demonstrate a new small spacecraft technology that aims to lower mission costs and expand access to space.

The launch window opens Dec 18 at 12am ET. Live coverage from @RocketLab begins approximately 20 minutes before launch.

Learn more: https://t.co/0q14uDXa6a

[Nitin Kaithvar @Nitinkaithvar]

@jacobrodri_ ⚡️

[Nitin Kaithvar @Nitinkaithvar]

Gemini 3 flash is here

[Google AI Studio @GoogleAIStudio]

⚡️ gemini 3 flash is here

• frontier intelligence built for speed
• pro-grade reasoning at a fraction of the cost
• our most impressive model for agentic workflows

available now as gemini-3-flash-preview in AI Studio and the Gemini API

[Nitin Kaithvar @Nitinkaithvar]

gpt image 1.5 just landed
faster, sharper, better edits
but does it actually beat nano banana pro?

[Nitin Kaithvar @Nitinkaithvar]

everyone’s screaming “nano banana killer” about gpt image 1.5
nah
it’s faster and cleaner than old gpt image 1 but nano banana pro still smokes it on text rendering and prompt adherence
openai caught up—not leaped ahead

[Nitin Kaithvar @Nitinkaithvar]

tested gpt image 1.5 against nano banana pro on the hardest prompts
nano banana still nails clean text in images every time
gpt 1.5 gets closer on realism but mangles letters half the runs
nano banana stays on top for now
change my mind

Guess which one is ChatGPT?

[Nitin Kaithvar @Nitinkaithvar]

Prompt: {
"subject": {
"demographics": "Young adult female, approx 20-25 years old, Caucasian.",
"hair": "Voluminous ash-blonde hair with honey highlights, deep side part, layered blowout style with significant root volume, cascading in thick waves over the left shoulder and back.",
"face": {
"shape": "Heart-shaped face with high cheekbones and a sharp, defined jawline.",
"eyes": "Almond-shaped blue-grey eyes, heavy winged eyeliner, long lashes, direct alluring gaze.",
"lips": "Full, plump lips with a nude-pink glossy finish and darker lip liner (overlined style).",
"skin": "Tan complexion, smooth but textured, visible contouring makeup."
},
"body_proportions": {
"build": "Voluptuous hourglass figure, slim waist contrasting with wide hips and a prominent bust.",
"chest": "Full, heavy bust volume visibly filling the top, deep cleavage, natural projection and gravitational weight, tight fit.",
"anatomy": "Natural soft-tissue behavior, no artificial flattening, realistic chest-to-waist ratio preserved."
},
"attire": {
"top": "White ribbed cotton tank top, deep scoop neckline, tight form-fitting silhouette, thin straps.",
"bottom": "Olive green/khaki cargo pants or denim, visible waistband, casual fit."
}
},
"pose": {
"orientation": "Seated, leaning diagonally back and to the right.",
"limbs": {
"right_arm": "Extended forward towards the camera lens (out of frame), implying a handheld selfie angle.",
"left_arm": "Extended back, hand likely resting on the surface behind for support, shoulder dropped slightly.",
"legs": "Angled to the left, knees bent, seated position."
},
"head": "Tilted slightly to the left, chin down slightly, looking up through eyelashes.",
"dynamics": "Relaxed but poised, upper body slightly twisted relative to hips."
},
"setting": {
"environment": "Bright, modern interior living space.",
"background_elements": "White sectional sofa with plush texture, a patterned beige pillow on the left, a window in the upper background letting in daylight.",
"surfaces": "White fabric of the couch/bed she is sitting on.",
"atmosphere": "Casual, domestic, daytime luxury."
},
"camera": {
"shot_type": "High-angle close-up (POV/Selfie perspective but without visible phone/mirror).",
"framing": "Upper body and thighs, subject centered but angled diagonally.",
"focal_length": "24mm to 35mm (slight wide-angle distortion typical of handheld shots).",
"perspective": "Looking down at the subject, emphasizing the eyes and upper body volume.",
"depth_of_field": "Shallow depth of field, face and torso in sharp focus, background sofa and window slightly soft.",
"aspect_ratio": "3:4"
},
"lighting": {
"source": "Strong natural daylight coming from a window on the left.",
"quality": "High contrast sunlight mixing with ambient fill.",
"direction": "Side-lighting creating highlights on the right side of the hair and left side of the chest.",
"shadows": "Soft shadows on the right side of the face and neck, defining the jawline and collarbones.",
"highlights": "Bright specular highlights on the blonde hair, lip gloss, and shoulders."
},
"mood_and_expression": {
"emotion": "Confident, sultry, relaxed.",
"micro_expressions": "Slight pout, intense eye contact, calm demeanor.",
"vibe": "Instagram influencer aesthetic, casual chic."
},
"style_and_realism": {
"style": "Hyper-realistic raw photography.",
"fidelity": "High fidelity, unpolished texture.",
"details": "Visible fabric weave on the tank top, individual hair strands, skin pores, natural skin variation (moles/freckles if applicable, makeup texture).",
"imperfections": "Natural fabric folds on the pants and top, realistic hair flyaways."
},
"colors_and_tone": {
"palette": "Neutral and earthy tones. White (top/sofa), Olive Green (pants), Ash Blonde (hair), Tan (skin).",
"contrast": "Medium-high contrast due to direct sunlight.",
"saturation": "Natural saturation, slightly warm skin tones.",
"white_balance": "Daylight balanced, slightly cool in shadows, warm in highlights."
},
"quality_and_technical": {
"resolution": "8k",
"sharpness": "Razor sharp on eyes and lips.",
"noise": "Low ISO, minimal grain.",
"texture_fidelity": "High."
}
}

[Nitin Kaithvar @Nitinkaithvar]

openai just dropped gpt image 1.5
people are already calling it the nano banana killer
i ran the same prompts on both
results coming up

GPT Image 1.5 vs Nanobanana Pro

Guess which one is GPT Image 1.5?

[Nitin Kaithvar @Nitinkaithvar]

Bro what Avengers Doomsday trailer got leaked?

[Nitin Kaithvar @Nitinkaithvar]

Which AI you use most?

[Nitin Kaithvar @Nitinkaithvar]

Bro what?
We got GTA 7 Trailer before GTA 6 😅

[Nitin Kaithvar @Nitinkaithvar]

Figma going all-in on AI image generation powered by Gemini 3. The design tooling landscape is shifting fast. This direct competition with Photoshop using cutting-edge AI capabilities shows how serious design tools are getting about AI integration. 🔄

[el.cine @EHuanglu]

Figma just launched AI image editing tool.. to replace Photoshop

powered by Gemini 3 with Nano Banana

its so over for traditional softwares

[Nitin Kaithvar @Nitinkaithvar]

Garlic code red moment 🚨 OpenAI's strategy shift got real. Google crushing them in benchmarks while ChatGPT usage drops. Sam Altman pausing projects speaks volumes. This is the AI competitive landscape being redrawn in real-time. The market's moving faster than anyone predicted. 🔬

[Julian Goldie SEO @JulianGoldieSEO]

OpenAI just leaked a model so powerful it triggered code red.

And yes, it’s called Garlic.

Here’s the 30-second breakdown so you actually understand why this matters →

✔ Google Gemini 3 crushed OpenAI in benchmarks
✔ ChatGPT usage dropped
✔ Sam Altman paused projects to build “Garlic”
✔ Garlic uses high-density pre-training (smarter + smaller)
✔ Internal memos show it beating Gemini 3 + Claude Opus 4.5

Why this is wild:

→ Faster than current models
→ Cheaper to run
→ Better at reasoning
→ Better at coding
→ Could run locally on your phone

Save this video, you'll rethink what AI models can do.

Want the SOP? DM me. 💬

[Nitin Kaithvar @Nitinkaithvar]

"Chain of debate" - I love this approach. Multiple models debating through a framework to reach better decisions. This is how advanced reasoning should work: diverse perspectives, structured thinking, real-time synthesis. The future of AI decision-making is here. 🚀

[Satya Nadella @satyanadella]

Was fun to be at a dev event in Bengaluru and demo an app I built recently for deep research with multiple models and decision frameworks...think of it as "chain of debate"... Next stop, Copilot!!