Image to Video AI Free [2026]: Kling 4.0 vs Veo 3.1

Free image to video AI is thinner in November 2026 than the search results suggest. Three models carry most serious work: Kling 4.0 from Kuaishou, Runway Gen-4.5, and Google’s Veo 3.1. OpenAI’s Sora, the fourth name that anchored most comparisons a year ago, no longer exists as a service. Runway’s Gen-4.5 needs a paid plan before you can generate a single frame.

The numbers that matter most for animating a still come down to clip length and cost. Kling 4.0 launched on October 15, 2026 with native clips of 3 to 30 seconds, up to 10 keyframes and up to 15 reference assets. Google’s Veo 3.1 page positions one model for text-to-video, image-to-video and generated audio. Runway charges 12 credits per second for Gen-4.5.

This comparison puts the three side by side on specs, free access, price and fit for real jobs such as channel intros and logo reveals. It is narrower than our ranked list of seven image-to-video tools, which covers more models. Every figure below comes from a vendor page or release note read on November 2, 2026, and anything I could not confirm is labelled that way.

Image to video AI free: the short answer

Kling 4.0 is the pick for long clips, Veo 3.1 is the pick for short vertical clips inside Google’s apps, and Runway Gen-4.5 is the pick when you want clear credit math and several models behind one login. None of the three is free in the sense of unlimited and watermark-free. That is the honest summary of a Kling vs Runway vs Veo comparison in late 2026.

Here is how the three line up on the points a creator checks first:

• Longest clip: Kling 4.0 at a native 30 seconds. Google lists 4, 6 or 8 seconds for Veo 3.1 Lite, and I found no longer ceiling for the full model in the sources.
• Most reference control: Kling 4.0 with up to 15 reference assets and 10 keyframes. Veo 3.1 answers with Ingredients to Video, First and Last Frame and Scene Extension.
• Clearest price: Runway, which publishes Gen-4.5 at 12 credits per second and shows how many seconds each plan buys.
• Free access: Gen-4.5 requires the Standard plan or higher. Kling lists a free Basic tier. Google Vids has offered a small monthly allowance of free Veo 3.1 clips on personal accounts.
• Retired: Sora. The app and website ended on April 26, 2026, and the API ended on September 24, 2026.

I judge each model by its output, not its launch copy. That means clip length, resolution, watermark, cost, and the failure modes a vendor admits to. Where a vendor has not published a number, the tables below say “not confirmed” instead of filling the gap with a guess.

The rest of the article follows the order a creator usually works in. First the three models one at a time, then a 12-row specs table, then free tiers, pricing, published benchmarks, real projects, use-case picks, render settings, a migration guide and a verdict.

Sora is gone, and who fills the gap

If an older guide tells you to animate a still with Sora 2, it is out of date. OpenAI’s own video documentation says the Sora 2 models and the Videos API were shut down on September 24, 2026. A separate OpenAI page states plainly: “As of April 26, 2026, the Sora product is no longer available.” Two dates, two shutdowns: the consumer web and app experience first, the developer API five months later.

For creators, the practical impact is on any workflow built around the sora-2 or sora-2-pro model names. Those calls no longer work. I found no official source showing a renamed or acquired successor, so treat reports of a Sora spin-off as rumor until OpenAI says otherwise. Our ranking of AI video generators already reflects the change, and the migration section later in this piece covers the move step by step.

The models that took over

Most Sora users landed on the same short list. Beyond the big three, two other makers keep a current model on their own sites. Luma’s homepage presents Luma Ray 3.2 as its newest line, and Pika’s site presents Pika 2.5. I could not confirm clip lengths or free-tier terms for either from those pages, so this article does not rank them on specs.

Runway’s developer platform also opens a back door to open and partner models. Runway Dev lists Wan 3.0 and Seedance 2.0 alongside its own Gen-4.5, all billed from one credit balance. Runway’s pricing page also advertises Seedance 2.5 as a multi-reference model for clips up to 30 seconds, though I did not find a price for it. The per-second rates for Wan 3.0 and Seedance 2.0 appear in the pricing section.

What about Veo 4?

Nothing official exists. A public third-party model list contains a “Veo 4” entry with no price, but Google’s own pages still list Veo 3.1 as the newest Veo model as of this writing. Treat Veo 4 as a rumor. Google’s recent activity has been about spreading Veo 3.1 into more products, not releasing a successor.

Kling 4.0: the long-clip model

Kling 4.0 is the longest-running option on this list, with native generation from 3 to 30 seconds. Kuaishou, the parent company, announced it in a release dated October 16 in Hong Kong and published on October 17, and Kling’s own news page dates the launch to October 15, 2026. Kling describes it as a next-generation model with improved motion and camera continuity, Omni Reference support for up to 15 reference assets, and up to 10 keyframes.

For image-to-video work, the keyframes and reference slots are what set it apart. A keyframe pins the video to a specific still at a specific moment. Give the model a start image and an end image and it has to bridge them, which is the safest way to control a logo reveal or a product turn. Fifteen reference assets let you hold a character, a prop and a background steady across shots instead of hoping one image is enough.

Where Kling 4.0 sits in the lineup

Kling 4.0 arrived after a fast run of releases. The Kling 3.0 family launched in February 2026, and Kuaishou’s second-quarter results on August 19 reported a Kling 3.0 Turbo release and native 4K output rolling out across the 3.0 series. A third-party tracker also reports a Kling 4.0 Flash tier in limited early access from September 28. I could not confirm native 4K for Kling 4.0 itself, so check the resolution picker in the app before promising a client a 4K deliverable.

Access is simple on the consumer side. Kling 4.0 is available to all users through the Kling AI product. The developer side is slower. Reporting around launch described API access as early access with no official rate card, so you cannot price an automated pipeline from the sources I found. The KlingAI Open Platform documents the API for earlier models, and it is the place to watch for 4.0 terms.

Plans, credits and the free tier

Kling’s plan names are Standard, Pro, Premier and Ultra+, according to a third-party review. The same tracker lists Standard at $8.80 a month for 660 credits and Pro at $32.56 for 3,000 credits as of October 29, while an earlier review quoted $10 and $37 for the same two tiers. Kling’s own pricing page warns that prices, credits and access rules may change, and it showed a 10% Kling 4.0 discount in October. Use the lower, more recent figure as a guide and confirm at checkout.

Kling’s pricing page reportedly shows the free Basic plan with no monthly credits and marks generated content as not for commercial use. The Standard plan lists 1080p generation, watermark removal, video extension and commercial rights. For a monetized YouTube channel, that split matters. A free-tier clip is fine for a test but not for the intro you plan to run on every upload.

What Kling admits it struggles with

The vendor is candid in one place. In Kling’s own case studies, director Kim Sung-soo resumed his AI short film WormTrain after Kling 4.0 produced physical collisions that earlier models struggled to generate. That is a useful signal about the failure mode of the previous generation, and also a reminder that collisions and contact between objects remain the hard cases for any model in this class.

Runway Gen-4.5: paid-only and priced in credits

Runway Gen-4.5 is Runway’s own flagship video model, and it is not free. Runway’s help article on creating with Gen-4.5 and its pricing page describe it as supporting text-to-video and image-to-video, requiring the Standard plan or higher, and costing 12 credits per second. I found no evidence of a newer Runway-native flagship released after Gen-4.5.

The pricing page makes the math easy. The entry plan’s 625 monthly credits buy 52 seconds of Gen-4.5 or 104 seconds of Gen-4 Turbo. The next tier up, at 2,250 credits, buys 187 seconds of Gen-4.5 or 375 seconds of Gen-4 Turbo. The top tier, at 9,500 credits, buys 791 seconds of Gen-4.5 or 1,900 seconds of Gen-4 Turbo. Gen-4 Turbo is the cheaper draft model, at half the credit cost per second by those figures.

Runway as a multi-model platform

Runway now does two jobs. It is a creative app with its own models, and it is a developer platform that resells other labs’ models under one key and one credit balance. Runway Dev exposes Wan 3.0, Seedance 2.0 and Kling models, including Kling 4.0, next to Gen-4.5. That has a consequence for any Kling vs Runway test: you must say whether “Runway” meant Gen-4.5 or Kling 4.0 running inside Runway. Those are different models with different prices.

A Runway Dev guide updated October 22, 2026 says new models ship most weeks. That cadence is good news if you like choice and bad news if you need a stable pipeline, because a model you pick today may sit beside a better one by next month. Runway’s pricing page also carries a promotion at the time of writing: 25% off Kling 4.0 generations until November 30.

The Unlimited-to-Max change

Subscribers on the old Unlimited plan should note the date. Runway introduced a Max plan on June 1, 2026 and began phasing out Unlimited. According to Runway’s help centre, Unlimited gave 2,250 credits plus Explore Mode, and on November 30 monthly Unlimited subscribers are scheduled to move automatically to Max at the same $95 a month unless they cancel. If you run a channel on an Unlimited plan, check what Explore Mode behavior you lose before that date. For more on this topic, see our image-to-video guides.

Veo 3.1: short clips, vertical output, Google’s apps

Veo 3.1 is Google’s current video model, and the only one of the three that most people can already reach inside tools they use daily. Google says it is available in the Gemini app, Flow, Google Vids, YouTube Shorts, the Gemini API and Vertex AI. Google DeepMind’s page calls it a new era for video generation and lists text-to-video, image-to-video and text-to-audio plus video generation.

The most recent major update improved Ingredients to Video, which builds a clip from reference images and aims to keep a character and background consistent. It also added native vertical 9:16 output, improved 1080p and 4K upscaling. Google also brought the same Ingredients to Video and portrait mode to YouTube Shorts and the YouTube Create app. For a creator who publishes Shorts, a vertical-native model avoids the crop step that wastes pixels.

Clip length and the Lite model

Google’s launch post for Veo 3.1 Lite, dated March 31, 2026, describes it as the most cost-effective Veo model. It supports text-to-video and image-to-video, 16:9 and 9:16, 720p and 1080p, and 4, 6 or 8 second clips. Google also said Lite would cost less than half of Veo 3.1 Fast and that Fast pricing would drop on April 7. I could not confirm a longer clip length for the full Veo 3.1 model, so plan around eight seconds as the working ceiling.

Eight seconds is enough for a title sting or a logo reveal. It is short for anything that needs a story beat. Veo offers extension tools, including Scene Extension and First and Last Frame, but each added segment costs another generation, and my sources give no per-clip price to estimate that.

Control features beyond the image

Google’s Veo page lists four named capabilities: Ingredients to Video, Scene Extension, First and Last Frame and Object Insertion. The page says these reached leading results in head-to-head comparisons by human raters on internal benchmarks. That is Google grading its own work, and the page was last updated in October 2025, so it predates the Kling 4.0 launch. Read it as a statement about Veo’s strengths, not as proof it beats Kling today.

Specs table: Kling 4.0 vs Runway Gen-4.5 vs Veo 3.1

The table below uses only figures from the vendors’ own pages and release notes, plus two third-party price trackers where marked. Cells that say “not confirmed” mean I found no figure I could trust for that model. Check the vendor’s current page before you commit to a project that depends on one of them.

Two rows deserve a comment. The clip-length gap is the biggest practical difference, with Kling’s 30 seconds against Veo’s 8. The “not confirmed” cells for Runway on length, ratio and resolution come from a documentation gap in what I could read, not from a real limit, since Runway’s app clearly offers a duration picker.

Free tiers compared: what you get before you pay

For anyone searching for an image to video AI free of charge, the picture splits three ways. Runway gives nothing free for Gen-4.5, because the model requires the Standard plan or higher. Kling lists a free Basic plan. Google has let personal accounts try Veo 3.1 for free in Google Vids. Each of those facts comes with a date and a catch, so read the details before you plan a video around them.

Kling Basic

According to Kling’s pricing page, the Basic plan is reportedly free forever, shows no monthly credits and marks generated content as not for commercial use. A pricing page from October 6 adds that credits, benefits and access rules may change as Kling updates. In plain terms, a free Kling account is a way to see how the model handles your image. It is not a way to produce a channel intro you can monetize.

The same snapshot lists the Standard plan with 660 monthly credits, 1080p generation, image upscaling, watermark removal, video extension and commercial-use rights. The watermark-removal line implies free output carries a mark, though I did not find the exact watermark rules on a page I could read. Check the first export before you build anything on it.

Veo 3.1 in Google Vids

Google reportedly said that personal Google accounts could generate 10 Veo 3.1 videos a month at no cost in Google Vids, and that Google AI Ultra and Workspace AI Ultra accounts could generate up to 1,000 Veo videos a month there. Google changes limits often, so confirm the current allowance inside Vids. According to Google’s consumer plan page as retrieved in late October, the lineup is Google AI Plus, AI Pro and AI Ultra. That page does not clearly state a separate Veo allowance, and I did not find a confirmed AI Pro price.

For a one-off test of a photo to video AI free workflow, Vids is the lowest-friction route if you already have a Google account. You get Veo-quality motion on a short clip without entering a card. The cost is control, since I saw no sign of keyframe or multi-shot tools in Google’s announcement.

Runway

Runway’s pricing page is headed “from $12/month” and Gen-4.5 sits behind the Standard plan. The cheapest way to try Runway’s own model is therefore a paid month on the entry plan, where 625 credits give 52 seconds of Gen-4.5. If you only want to test image animation, Gen-4 Turbo stretches the same credits to 104 seconds, which suits quick tests of how a still will move.

What free output means for a channel

A practical rule for creators: use free tiers to audition, then pay for the final render. Free output tends to carry marks, tighter limits and, in Kling’s case, a non-commercial label. If you want an intro without spending credits at all, an original theme from a template library is a cheaper path. The intro themes list resolution, length and frame rate for each design, which makes them a useful benchmark for what your AI-made clip should match.

Pricing: subscriptions and per-second rates

Pricing is where these tools differ most in how clearly they explain themselves, and it decides how far an image to video AI free trial can take you. Runway shows credits and seconds. Kling shows credits and plan names but changes promotions often. Google ties Veo to broader plans and to developer pricing I could not read in full. The two tables below keep to figures I could verify.

Runway plans: credits and seconds of Gen-4.5

Divide the entry plan’s $12 by 52 seconds and a second of Gen-4.5 costs about 23 cents, a number you can compare against the API rate below. The monthly price without annual billing is higher, so the real figure at monthly billing will be worse. My math, not Runway’s.

Per-second rates on Runway Dev

Runway Dev states that one credit equals one cent for its documented API pricing. On that basis, Gen-4.5’s 12 credits a second is $0.12 per second, and a five-second clip is 60 credits, or $0.60. The table adds the partner models that Runway Dev lists, since they are the cheapest ways to draft a clip.

A cheap draft at 480p on Wan 3.0 Prime costs $0.068 a second, so a ten-second test is 68 cents. The same ten seconds of Seedance 2.0 at 1080p is $4.00. Those gaps show why a draft-then-final workflow saves money: use the cheap model to settle composition, then spend credits on the model you plan to deliver.

Kling and Google

Kling’s consumer price points are the third-party figures noted earlier. On the developer side, I found no official Kling 4.0 rate card, and Runway lists Kling 4.0 at its own credit price with the 25% promotion. For Google, Veo 3.1 Lite is reportedly priced under half of Veo 3.1 Fast according to Google’s March 31 announcement, and I did not find a clean per-second figure for the full model. If you plan to build on the Gemini API or Vertex AI, read Google’s current developer pricing before you budget.

Benchmarks: what the published tests show

There is no independent head-to-head benchmark of Kling 4.0, Gen-4.5 and Veo 3.1 on image-to-video that I could confirm. What exists are three kinds of vendor-published evidence, each from a different source, and each limited in a specific way. Treat them as inputs to a decision, not a scoreboard.

1. Google DeepMind’s rater studies. The Veo page reports human preference results on public benchmark sets. For image-to-video, participants viewed 355 image and text pairs from VBench I2V and preferred Veo 3.1 for capturing the prompt’s intent and for visual quality. On MovieGenBench, a Meta dataset, 1,003 prompts were used for overall preference and prompt following, and 527 prompts covered audio sync. The limit: the page was last updated in October 2025 and compares Veo to unnamed “other models,” so it cannot speak to Kling 4.0.
2. Kuaishou’s Kling 4.0 release. Kuaishou describes more stable high-dynamic motion, complex camera movement, native clips up to 30 seconds, 15 reference assets and 10 keyframes. These are capability claims, not scored comparisons. Kling’s blog adds creator reports, including one director who found most Kling 4.0 test footage usable in editing or kept as alternate takes.
3. Runway’s credit throughput. The pricing page converts credits to seconds: 52, 187 and 791 seconds of Gen-4.5 across the three tiers, doubled for Gen-4 Turbo. This is a cost benchmark, not a quality one, but it is the only one of the three that you can verify with a calculator.

A fourth data point comes from Kuaishou’s second-quarter 2026 results, reported on August 19. They credit the Kling 3.0 series with native 4K output and a Kling 3.0 Turbo release aimed at faster production and audio-visual synchronization. It confirms the direction of travel, not a measured gain.

How to run your own test

Because the public numbers do not settle the question, run a small bake-off. Pick one still, such as your logo on a plain background, and one prompt. Generate the same five seconds on each model, then score on four counts: does the logo keep its shape, does the camera move as asked, is there a watermark, and what did the clip cost in credits. Ten minutes and a few dollars beat a week of reading comparison pages. If one of the three fails the logo-shape check, drop it for that job.

Real-world examples from studios and creators

All of the following come from vendor pages, so they are vendor-reported. They show what each platform wants to be known for, which is useful on its own.

• An April Fools’ Day ad for The RealReal. Kling’s showcase lists a spot by director Sebastian Strasser and his studio Lipstick for the luxury resale platform. It won one Silver and one Bronze at the 2026 Cannes Lions International Festival of Creativity.
• An ad from Purga Films. The Argentine studio promoted its AI advertising services with a spot made entirely with Kling AI, which took Bronze in the Classic Film category at Cannes Lions 2026.
• A viral short film. Kling’s page describes an AI short that circulated during the 2026 Qingming Festival, made entirely with Kling AI. It reportedly passed 100 million views across platforms.
• A 135-minute feature. The AI studio Fountain 0 and Kling AI co-produced a feature directed by Ash Koosha, and Kling says all 135 minutes of its visuals came from Kling.
• A fashion commercial. Tom-Jacques Perret, co-founder of UNVEIL, used Kling 4.0 while developing a commercial for the Danish brand Heliot Emil.
• Previs on Veo 3.1. Google’s page says Promise Studios uses Veo 3.1 inside its MUSE Platform for generative storyboarding and previsualization.
• Game cinematics on Veo 3.1. Volley uses Veo 3.1 in its RPG Wit’s End for static cinematics and dynamically generated assets that narrate player progress.
• Promo videos on Veo 3.1. OpusClip uses it inside Agent Opus to build motion graphics and promotional videos for small and mid-sized businesses.
• Ad recipes on Runway. Runway Dev documents “campaign-ready ads and images” in one API call. Its listed recipe costs run from 9 credits for a low-quality stock image to 188 credits for a four-second 720p Product Swap.

What do these examples tell a solo creator? Kling’s showcase leans on long, finished work, from ads to a feature film. Google’s leans on tools: storyboards, cinematics and promo clips inside other products. Runway’s leans on pipelines and cost per recipe. Pick the platform whose public story looks like the job you do.

Best model by use case

The right pick depends on what the clip is for. Here are seven common jobs, each with a primary pick and a runner-up, based on the specs and prices above.

1. YouTube channel intro, 5 to 8 seconds. Veo 3.1 or Kling 4.0. Either handles the length, and Kling’s keyframes give tighter control if you start and end on a logo card. Veo 3.1 is the lower-friction pick if you only have a still and a prompt.
2. Logo reveal from a flat PNG. Kling 4.0 with start and end keyframes. Pin the first frame to a blank background and the last frame to the finished logo, and the model has less room to distort the mark. If you would rather start from a designed animation, the animated logo intro theme is an original design to compare against.
3. Vertical Shorts and Reels loops. Veo 3.1, which outputs native 9:16 and sits inside YouTube Shorts and YouTube Create. Runner-up: Kling, once you confirm its vertical option.
4. Product ad with consistent branding. Runway, because its credit math is explicit and it lists campaign recipes with costs. Kling 4.0 through Runway is a close second while the 25% promotion lasts, until November 30.
5. Long B-roll or music-video scenes. Kling 4.0 at up to 30 seconds per native clip. No other model in this comparison lists a native length near that.
6. Storyboards and previs. Veo 3.1, going by Promise Studios’ use of it for generative storyboarding.
7. Cheap drafts. Wan 3.0 Prime on Runway Dev at $0.068 a second for 480p. Settle composition there before paying for a final model.

No single model wins all seven, and a free image to video AI account on one of them rarely covers more than the first test. A creator who makes weekly videos will often use two: a cheap draft model and a finishing model. Switching is easier than it sounds, since Runway Dev puts several models behind one balance.

Render settings for intros and logo reveals

Whatever model you pick, the clip still has to survive an editor and an upload. These are my suggested targets for the common jobs, not vendor specs. They assume you generate at the highest resolution your plan offers, then conform the clip to your timeline.

Three habits cut re-renders. First, start from a clean source image at the target aspect ratio, because a model that has to guess at cropped edges tends to invent content there. Second, put the logo on a plain, flat background so any warping is easy to spot and mask. Third, check the first and last frames of every clip before you build a sequence, since those frames decide how well your cuts match.

An AI clip often arrives at a frame rate that differs from your timeline. Set your editor to the frame rate of the project, not the clip, and let the clip conform. If you notice stutter, retime the clip with optical flow rather than regenerate it. Regenerating costs credits, while retiming costs seconds.

Migration guide: moving a workflow between models

A change of model is a routine event now. Runway adds models most weeks, Sora’s shutdown forced a wave of moves, and Kling went from the 3.0 family in February to 4.0 in October. Use this sequence to move without losing work.

1. Export what you still hold. If you made clips in Sora, check OpenAI’s shutdown notice for the export window and save every file, prompt and source image first.
2. Inventory your prompts and stills. Keep a spreadsheet with the source image, the prompt, the duration and the credits spent per clip. You need this to price the move.
3. Map each job to a model. Use the use-case section above. Long scenes go to Kling 4.0, vertical loops to Veo 3.1, and ad recipes to Runway.
4. Re-test one clip per job. Never migrate a whole library at once. Run the same still and prompt on the new model and compare logo shape, camera move, watermark and cost.
5. Rewrite prompts for the new control scheme. A prompt tuned for a text-only model rarely transfers. With Kling, move detail from the prompt into keyframes and reference assets. With Veo, use Ingredients to Video to hold characters steady.
6. Check the plan terms. Confirm commercial rights and watermark rules on the plan you will use. Kling’s free tier is marked non-commercial, and Runway’s Gen-4.5 needs the Standard plan.
7. Watch the dates. Runway’s 25% Kling 4.0 promotion ends November 30, and monthly Unlimited subscribers move to Max the same day. Time a switch around both.
8. Keep the old files. Store the original clips and metadata next to the new ones for at least one quarter, in case a model update changes behavior.

If you run a Runway Dev integration, the move between models is a configuration change, since the documented models share one key and one credit balance. If you call a vendor’s API directly, expect to rewrite the request format as well.

Pros and cons

Each model’s strengths line up with the facts above. The lists are short on purpose.

Kling 4.0

• Pros: native clips up to 30 seconds, up to 10 keyframes, up to 15 reference assets, listed free Basic tier, commercial rights on paid plans.
• Cons: no official 4.0 API rate card yet, plan prices differ across trackers and change with promotions, free output is marked non-commercial.

Runway Gen-4.5

• Pros: clear credit-to-seconds math, several partner models under one balance, a cheaper Gen-4 Turbo for drafts, documented Enterprise agreements that include IP indemnification.
• Cons: no free access to Gen-4.5, no newer Runway-native model confirmed since launch, the Unlimited plan is ending, and I found no firm clip-length or resolution figures for Gen-4.5.

Veo 3.1

• Pros: native 9:16 output, reference-image control through Ingredients to Video, generated audio, reach across Gemini, Flow, Google Vids, YouTube Shorts and two developer platforms.
• Cons: short clips (4 to 8 seconds in the Lite listing), benchmark claims that predate Kling 4.0, and no clear separate Veo price on Google’s consumer plan page.

Verdict: which one to pick

For most creators who want to animate a still this month, the choice follows the length of the clip. Pick Kling 4.0 if the clip runs past eight seconds or you need to pin start and end frames. Pick Veo 3.1 if the clip is short, vertical, and you already live in Google’s tools. Pick Runway Gen-4.5 if you want priced seconds and the option to switch between several models without a second login.

The data behind that call is short. Kling lists a 30-second native ceiling against Google’s 8 seconds for Veo 3.1 Lite. Runway’s entry plan buys 52 seconds of Gen-4.5 for a price that starts at $12 a month on annual billing, which works out to about 23 cents a second. One low-cost draft option on Runway Dev is Wan 3.0 Prime at $0.068 a second at 480p. On free access, the best route I found pairs Kling’s Basic tier for tests with Google Vids for a few Veo clips, as long as you accept the non-commercial label and the monthly cap.

Image to video AI free tools are best used as an audition stage. Run your still through the free allowances, pick the model that keeps your logo intact, then pay for the final render on a plan that grants commercial rights. Two caveats apply. First, none of the benchmark evidence is independent, so your own five-second bake-off carries more weight than any vendor chart. Second, the market moves quickly, with Runway adding models weekly and Kling moving from 3.0 to 4.0 inside nine months, so recheck prices and limits before each big project.

Questions creators ask

Is there a truly free image to video AI? In short, no. Image to video AI free access exists, but not in the sense of unlimited, watermark-free and commercial. Kling lists a free Basic plan marked non-commercial, and Google Vids has offered a small monthly allowance of Veo 3.1 clips on personal accounts. Runway’s Gen-4.5 requires the Standard plan or higher. Treat free tiers as a test stage.

Is Sora still available? No. OpenAI ended the Sora web and app experience on April 26, 2026, and shut down the Sora 2 models and the Videos API on September 24, 2026. I found no official successor or rename.

Kling vs Runway: which is better for animating a logo? Kling 4.0 gives you up to 10 keyframes, which lets you pin a blank start frame and a finished logo end frame. Runway Gen-4.5 supports image-to-video but I found no keyframe count for it. You can also run Kling 4.0 inside Runway, with a 25% promotion until November 30, 2026.

Does Veo 3.1 image to video support vertical video? Yes. Google’s latest update added native 9:16 output, and the Veo 3.1 Lite listing includes both 16:9 and 9:16 at 720p and 1080p.

Can I use free clips on a monetized YouTube channel? Check the plan terms first. Kling’s pricing page reportedly marks free-plan content as not for commercial use and lists commercial rights on paid plans. Rules differ by vendor and change often, so read the current terms before you publish.

How many seconds does Runway’s cheapest plan buy? The entry plan’s 625 monthly credits equal 52 seconds of Gen-4.5 or 104 seconds of Gen-4 Turbo. At the API rate of 12 credits a second, a five-second Gen-4.5 clip uses 60 credits.

Is there a Veo 4? Nothing official. A third-party list shows a Veo 4 entry with no price, but Google’s own pages list Veo 3.1 as the newest model. Treat Veo 4 as a rumor until Google announces it.