Fall 2026

Research Intern

Pika

Pika

51-200 employees

AI-powered video creation and editing tools

No salary listed

Palo Alto, CA, USA

Hybrid

Category
AI & Machine Learning (1)
Required Skills
Python
Tensorflow
Neural Networks
Pytorch
Machine Learning

People at Pika

People at Pika who can refer or advise you

Requirements
  • BS, MS, or PhD-level backgrounds; excellent PhD interns or exceptional undergraduates strongly encouraged to apply.
  • Experience or coursework in machine learning, deep learning, or multimodal AI.
  • Strong technical skillset—proficiency in Python and relevant ML frameworks (PyTorch, TensorFlow, etc.).
  • Quick learner able to contribute to fast-paced research projects and adapt to new technologies.
  • Excellent communication skills and ability to collaborate in cross-functional teams.
Responsibilities
  • Contribute directly to large-scale pre-training and mid-training efforts for multimodal foundation models.
  • Assist in designing, prototyping, and optimizing algorithms for high-throughput data pipelines and multimodal MLLM inference.
  • Work with diverse data types, including text, image, audio, and video, to support advanced generative model training.
  • Collaborate closely with experienced researchers and engineers on state-of-the-art core technologies.
  • Learn and apply frontier research techniques and production-scale AI workflows.
Desired Qualifications
  • Prior work or research in MLLM inference, generative models, or similar AI systems is a plus.

Pika.art builds an online platform that turns ideas into videos using AI. It offers text-to-video, image-to-video, and video-to-video transformations so users can create, edit, and extend videos with simple text commands. The product works through AI-powered video generation and editing tools accessible in a browser-like interface, enabling users to modify scenes, extend runtimes, and resize canvases without deep technical skills. Pika differentiates itself by targeting a broad range of creators—from casual meme makers to professional filmmakers—with a user-friendly, text-driven workflow and a subscription-based model that likely includes premium features for advanced tools. The company’s goal is to democratize video creation by making powerful AI-assisted video production accessible and customizable to everyday users and professionals alike.

Company Size

51-200

Company Stage

Series B

Total Funding

$135M

Headquarters

Palo Alto, California

Founded

2023

People at Pika

People at Pika who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • Pika 2.5, Pikaframes, and Pika Agent deepen control and workflow automation.
  • Predictive Video and selfie-driven features support emotionally expressive, social-first creation.
  • Freemium pricing and viral templates can convert mass consumer attention into paid usage.

What critics are saying

  • OpenAI Sora compresses Pika's differentiation in high-end generative video.
  • Runway and Luma raise user expectations for realism and cinematic quality.
  • Pika's short-clip format limits adoption for professional, long-form production workflows.

What makes Pika unique

  • Pika combines text-to-video, image-to-video, and video-editing in one creator-friendly product.
  • Pika targets everyday creators with no technical background, not just professional studios.
  • Pika monetizes through freemium subscriptions with tiered credit-based access and paid upgrades.

Help us improve and share your feedback! Did you find this helpful?

Benefits

Health Insurance

Company Equity

Growth & Insights and Company News

Headcount

6 month growth

1%

1 year growth

3%

2 year growth

10%
FreeLipSync
Jun 1st, 2026
FreeLipSync vs Runway vs Pika: which AI video tool should you actually use in 2026?

FreeLipSync vs Runway vs Pika: which AI video tool should you actually use in 2026? By Nina Brooks I've spent way too much time this year hopping between AI video tools, trying to figure out which one actually deserves a spot in my workflow. Runway and Pika get the most hype. But every time I compare them to FreeLipSync for lip sync work specifically, the answer keeps coming out the same. Let me break down exactly what each tool does, where it shines, and where it quietly lets you down. If you want cinematic video generation with camera motion and scene synthesis, Runway or Pika will impress you. If you want to make a photo or video speak - with any audio, in any language, at zero cost - FreeLipSync is the only tool that gives you that without a credit card or watermark. Before comparing prices, it's worth being honest about what each tool is built for. FreeLipSync is purpose-built for AI lip sync. You upload a face photo or video, provide audio (typed, uploaded, or voice-cloned), and it generates a realistic talking video. That's the entire use case - and it's extraordinarily good at it. Runway is a full-featured AI video studio. Gen-3 Alpha Turbo can generate entire video scenes from text prompts, extend clips, remove objects, apply motion effects. Lip sync is one feature among many, and it's geared toward professional post-production. Pika leans into character-driven animation. Its Pikaformance feature is impressive for turning images into expressive talking characters with emotional nuance. Great for creative content, less practical for corporate or educational use. These aren't the same tool. The question is which one matches your actual use case. | Tool | Free Tier | Cheapest Paid | | FreeLipSync | 20-sec videos, no watermark, no sign-up | $4.99/mo (Starter) | | Runway | 125 credits (~8 sec video), watermark | $12/mo (Standard) | | Pika | Limited free credits | ~$8/mo (Basic) | FreeLipSync's free tier is genuinely usable. No sign-up means you can generate your first video in under two minutes. No watermark means you can actually use it. The 20-second cap is a real limit, but for short social clips, announcements, or trying out a voice - it's plenty. Runway's free tier burns fast. At 125 credits, you're looking at maybe 8-10 seconds of generated video before you hit zero. Then it's $12/month minimum. Pika's Basic plan is affordable, but the free tier has become increasingly stingy with credits. If you're a solo creator trying to keep costs under $5/month, the decision is easy. I ran the same face photo through all three tools with identical audio. Here's what I found: FreeLipSync was the fastest - under 30 seconds - and the lip sync accuracy was genuinely impressive. The "Max" model adds more natural facial expressions and head movements. For a talking head video, this is production-quality. Runway produced a more cinematic result, but lip sync wasn't as tight. It's better suited to scripted video production where you're generating full scenes, not syncing a face to an existing voice. Pika (Pikaformance) delivered the most expressive character animation. If you're making a character talk with emotion - think animated storytelling - Pika is excellent. For a realistic human talking head, FreeLipSync wins on accuracy. FreeLipSync is right for you if: * You need to dub a photo or video in any language (500+ languages supported) * You want no sign-up friction and no watermark on short clips * You're a content creator, marketer, or educator making talking-head videos * Budget matters and you can't justify $12-30/month just for this one use case Runway is right for you if: * You need cinematic scene generation, not just lip sync * You're a video professional building full productions * You already pay for a creative suite and want AI embedded in a workflow Pika is right for you if: * You want expressive animated characters * You're making creative short-form content where emotional nuance matters more than strict realism I've seen a lot of "free AI tools" that are free in name only - watermarked, credit-capped, or gated behind a sign-up that feeds you into a sales funnel. FreeLipSync is genuinely different. The free tier is actually useful, which is rare. For most creators who need to make a talking video once a week or so, the free tier might be all you ever need. And if you need more - HD downloads, videos up to 60 minutes, voice cloning at scale - the Pro plan at $29.99/month is still substantially cheaper per video-second than Runway or HeyGen. Runway and Pika are impressive tools. But comparing them to FreeLipSync for lip sync work is like comparing a full recording studio to a great microphone. Both are good at what they do; only one is purpose-built for the thing you're actually trying to do. If you haven't tried FreeLipSync yet, go do it right now - no account needed: freelipsync.com

Daily Times Nigeria
Oct 19th, 2025
Pika AI App Raises $135M Funding

Demi Guo, a 27-year-old entrepreneur, launched the AI video app Pika in April 2023, targeting Gen Z users. After leaving her Ph.D. program at Stanford, Guo raised $135 million, valuing Pika at $470 million and attracting 16.4 million users. Pika's Predictive Video feature allows users to create videos from selfies, emphasizing personal expression. Competing with OpenAI's Sora and Meta's Vibes, Pika focuses on creativity and emotional storytelling for a digital-first audience.

Quasa
Jul 22nd, 2025
Pika Labs Launches TikTok-Style Social Network for AI-Generated Videos

Pika Labs launches tiktok-style social network for ai-generated videos.

Homebrew
Jun 30th, 2025
Pika Secures $55M in Funding

Pika, an AI-powered idea-to-video platform, announced it has raised $55 million in cumulative funding. The latest investment round was led by Mike Mignano, former CEO of Anchor and now a Partner at Lightspeed. Pika's CEO and Cofounder, Demi Guo, impressed investors with her technical achievements and focus on building a community.

Adweek
May 12th, 2025
Pika's 'Provocative and Controversial' Ad Casts AI as an Escape From a Post-apocalyptic World

About a year ago, Pika launched its 'Squish It' tool.