Key improvements:
*Extended duration:*\n\u2022 Up to 15 seconds of continuous video (vs previous 5-10 seconds)\n\u2022 Flexible duration ranging from 3-15 seconds\n\u2022 Better for complex action sequences and scene development
*Unified multimodal approach:*\n\u2022 Integrates text-to-video, image-to-video, reference-to-video\n\u2022 Video modification and transformation in one model\n\u2022 Native audio generation (synchronized with video)
*Two variants:*\n\u2022 VIDEO 3.0 (upgraded from 2.6)\n\u2022 VIDEO 3.0 Omni (upgraded from O1)
*Enhanced capabilities:*\n\u2022 Improved subject consistency with reference-based generation\n\u2022 Better prompt adherence and output stability\n\u2022 More flexibility in storyboarding and shot control
This positions Kling competitively against:\n- Runway Gen-4.5 ($95/month)\n- Sora 2 (limited access)\n- Veo 3.1 (Google)\n- Grok Imagine (just topped rankings)
The 15-second duration is particularly interesting - enables more narrative storytelling vs the typical 5-second clips. Combined with native audio, this could change workflows for content creators.
Pricing isn't mentioned in the announcement. Previous Kling models ranged from $10-40/month, significantly cheaper than Runway.
Anyone have access to test this yet? Curious how the quality compares to Runway and Sora at this new duration."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Kling VIDEO 3.0 released: 15-second AI video generation model"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://kling3.net"}},"_tags":["story","author_dallen97","story_46835895","show_hn"],"author":"dallen97","children":[46835970,46836020,46836099,46836337,46837337,46837615],"created_at":"2026-01-31T12:06:28Z","created_at_i":1769861188,"num_comments":2,"objectID":"46835895","points":4,"story_id":46835895,"story_text":"Kling just announced VIDEO 3.0 - a significant upgrade from their 2.6 and O1 models.
Key improvements:
*Extended duration:*\n\u2022 Up to 15 seconds of continuous video (vs previous 5-10 seconds)\n\u2022 Flexible duration ranging from 3-15 seconds\n\u2022 Better for complex action sequences and scene development
*Unified multimodal approach:*\n\u2022 Integrates text-to-video, image-to-video, reference-to-video\n\u2022 Video modification and transformation in one model\n\u2022 Native audio generation (synchronized with video)
*Two variants:*\n\u2022 VIDEO 3.0 (upgraded from 2.6)\n\u2022 VIDEO 3.0 Omni (upgraded from O1)
*Enhanced capabilities:*\n\u2022 Improved subject consistency with reference-based generation\n\u2022 Better prompt adherence and output stability\n\u2022 More flexibility in storyboarding and shot control
This positions Kling competitively against:\n- Runway Gen-4.5 ($95/month)\n- Sora 2 (limited access)\n- Veo 3.1 (Google)\n- Grok Imagine (just topped rankings)
The 15-second duration is particularly interesting - enables more narrative storytelling vs the typical 5-second clips. Combined with native audio, this could change workflows for content creators.
Pricing isn't mentioned in the announcement. Previous Kling models ranged from $10-40/month, significantly cheaper than Runway.
Anyone have access to test this yet? Curious how the quality compares to Runway and Sora at this new duration.","title":"Show HN: Kling VIDEO 3.0 released: 15-second AI video generation model","updated_at":"2026-03-05T23:27:50Z","url":"https://kling3.net"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"derek39576"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"Hi HN \u2013\nI hacked together AlephAI (https://alephai.net/) because I kept reaching for Runway\u2019s Gen-4 Aleph model but wanted a dead-simple, single-page UI I could put in front of non-technical teammates."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: AlephAI \u2013 Video-to-Video Style Transfer in the Browser"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://alephai.net"}},"_tags":["story","author_derek39576","story_44774299","show_hn"],"author":"derek39576","created_at":"2025-08-03T05:34:07Z","created_at_i":1754199247,"num_comments":0,"objectID":"44774299","points":1,"story_id":44774299,"story_text":"Hi HN \u2013\nI hacked together AlephAI (https://alephai.net/) because I kept reaching for Runway\u2019s Gen-4 Aleph model but wanted a dead-simple, single-page UI I could put in front of non-technical teammates.","title":"Show HN: AlephAI \u2013 Video-to-Video Style Transfer in the Browser","updated_at":"2025-08-03T05:35:09Z","url":"https://alephai.net"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Alisaqqt"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"Kling has been my go-to for I2V since the 2.0 days. The 3.0 release just dropped and it's a pretty significant jump, so I wanted to share what's new and how to actually get API access (including free credits to test it).
What's actually new in Kling 3.0?\nTwo models now \u2014 Kling 3.0 and Kling 3.0 Omni.
Biggest deal for me is the multi-shot thing. You describe a scene with dialogue and camera directions in one prompt, and it handles shot transitions automatically \u2014 close-up to wide, shot/reverse-shot, etc. I used to generate 3-4 separate clips and stitch them in Premiere. Now I get a coherent 10-15 second sequence in one pass. Not perfect every time, but when it works it saves hours.
Other stuff worth noting:
Native lip-sync across multiple languages (tested English and Spanish, both decent). No more bolting on a separate lipsync tool.\nText rendering is finally readable. Brand names on signs don't look like alphabet soup anymore.\nOmni lets you upload a short clip of someone and keep their appearance consistent across generations. Great for serialized content.
How does it compare?\nI've been testing across multiple platforms, so here's my honest take on where Kling 3.0 sits:
Where Kling 3.0 wins: Multi-shot storytelling in a single generation, multilingual lip-sync, and subject consistency (Omni). If you're doing anything with dialogue, characters, or non-English content, it's currently the strongest option.
Where others win: Veo 3.1 still has the edge on raw visual realism for single shots. Runway Gen-4's world consistency across separate generations is more mature. Sora 2 handles complex camera movements well.\nHow to get API access (step by step)\nKuaishou's direct API is a pain outside China. Region issues, payment issues, docs mostly in Chinese. Wasted a full afternoon on it.
Ended up using Atlas Cloud \u2014 basically an API aggregator, one key for Kling, Veo, Sora, Seedance, etc. Why I stuck with it:
- $1 free credit on signup (enough for a few test generations)\n- Playground to test prompts before writing code\n- Unified API, integration took ~10 minutes\nNot the only option, but it worked for me. If anyone has a better way to access Kling 3.0 API outside China lmk.
Pricing reality check
Atlas Cloud's Kling pricing runs about 30% below Kuaishou's direct API rates from what I can see. A few data points:\n- Kling 3.0 Standard: $0.126/sec\n- Kling 3.0 Pro: $0.168/sec\n- Kling O3 Standard: $0.07/sec\n- Kling O3 Pro: $0.168/sec
For comparison, Veo 3 through Google Cloud runs about $0.50/sec. So Kling is significantly cheaper for comparable (and in some cases better) quality.
Who is this actually useful for?\n- Short drama / serial content creators \u2014 The Omni subject consistency + multi-shot AI Director is basically built for this. Generate coherent multi-character scenes without stitching clips together.\n- Marketing / e-commerce teams \u2014 Multilingual lip-sync + accurate text rendering = localized video ads without reshooting.\n- Developers building video features \u2014 Standard REST API, OpenAI-compatible, no cold starts. Plug it into your pipeline.\n- Anyone comparing models \u2014 Having Kling, Veo, Sora, Seedance, etc. all behind one API makes A/B testing way less painful than juggling 5 different platforms.
TL;DR\nKling 3.0 is a legit step up \u2014 native audio-visual sync, 15s continuous generation, AI Director for multi-shot scenes, and strong subject consistency via Omni. You can access it through Atlas Cloud with free credits to test. The unified API approach means you can also compare it against Veo 3.1, Sora 2, etc. without separate accounts everywhere.
Happy to answer questions if anyone's tried it or wants to compare notes on specific use cases.\nLinks:\n- Atlas Cloud Kling collection: https://www.atlascloud.ai/models/list?q=Kling+3?utm_source=r...\n- Atlas Cloud playground: https://atlascloud.ai/models/explore?utm_source=reddit
Thanks."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"[dead]"}},"_tags":["comment","author_Alisaqqt","story_47015765"],"author":"Alisaqqt","comment_text":"Kling has been my go-to for I2V since the 2.0 days. The 3.0 release just dropped and it's a pretty significant jump, so I wanted to share what's new and how to actually get API access (including free credits to test it).
What's actually new in Kling 3.0?\nTwo models now \u2014 Kling 3.0 and Kling 3.0 Omni.
Biggest deal for me is the multi-shot thing. You describe a scene with dialogue and camera directions in one prompt, and it handles shot transitions automatically \u2014 close-up to wide, shot/reverse-shot, etc. I used to generate 3-4 separate clips and stitch them in Premiere. Now I get a coherent 10-15 second sequence in one pass. Not perfect every time, but when it works it saves hours.
Other stuff worth noting:
Native lip-sync across multiple languages (tested English and Spanish, both decent). No more bolting on a separate lipsync tool.\nText rendering is finally readable. Brand names on signs don't look like alphabet soup anymore.\nOmni lets you upload a short clip of someone and keep their appearance consistent across generations. Great for serialized content.
How does it compare?\nI've been testing across multiple platforms, so here's my honest take on where Kling 3.0 sits:
Where Kling 3.0 wins: Multi-shot storytelling in a single generation, multilingual lip-sync, and subject consistency (Omni). If you're doing anything with dialogue, characters, or non-English content, it's currently the strongest option.
Where others win: Veo 3.1 still has the edge on raw visual realism for single shots. Runway Gen-4's world consistency across separate generations is more mature. Sora 2 handles complex camera movements well.\nHow to get API access (step by step)\nKuaishou's direct API is a pain outside China. Region issues, payment issues, docs mostly in Chinese. Wasted a full afternoon on it.
Ended up using Atlas Cloud \u2014 basically an API aggregator, one key for Kling, Veo, Sora, Seedance, etc. Why I stuck with it:
- $1 free credit on signup (enough for a few test generations)\n- Playground to test prompts before writing code\n- Unified API, integration took ~10 minutes\nNot the only option, but it worked for me. If anyone has a better way to access Kling 3.0 API outside China lmk.
Pricing reality check
Atlas Cloud's Kling pricing runs about 30% below Kuaishou's direct API rates from what I can see. A few data points:\n- Kling 3.0 Standard: $0.126/sec\n- Kling 3.0 Pro: $0.168/sec\n- Kling O3 Standard: $0.07/sec\n- Kling O3 Pro: $0.168/sec
For comparison, Veo 3 through Google Cloud runs about $0.50/sec. So Kling is significantly cheaper for comparable (and in some cases better) quality.
Who is this actually useful for?\n- Short drama / serial content creators \u2014 The Omni subject consistency + multi-shot AI Director is basically built for this. Generate coherent multi-character scenes without stitching clips together.\n- Marketing / e-commerce teams \u2014 Multilingual lip-sync + accurate text rendering = localized video ads without reshooting.\n- Developers building video features \u2014 Standard REST API, OpenAI-compatible, no cold starts. Plug it into your pipeline.\n- Anyone comparing models \u2014 Having Kling, Veo, Sora, Seedance, etc. all behind one API makes A/B testing way less painful than juggling 5 different platforms.
TL;DR\nKling 3.0 is a legit step up \u2014 native audio-visual sync, 15s continuous generation, AI Director for multi-shot scenes, and strong subject consistency via Omni. You can access it through Atlas Cloud with free credits to test. The unified API approach means you can also compare it against Veo 3.1, Sora 2, etc. without separate accounts everywhere.
Happy to answer questions if anyone's tried it or wants to compare notes on specific use cases.\nLinks:\n- Atlas Cloud Kling collection: https://www.atlascloud.ai/models/list?q=Kling+3?utm_source=r...\n- Atlas Cloud playground: https://atlascloud.ai/models/explore?utm_source=reddit
Thanks.","created_at":"2026-02-14T16:26:04Z","created_at_i":1771086364,"objectID":"47015766","parent_id":47015765,"story_id":47015765,"story_title":"[dead]","updated_at":"2026-03-05T23:32:39Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"xnx"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"This is press release spam. Better to link to the announcement: https://runwayml.com/research/introducing-runway-gen-4.5"},"story_title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["runway"],"value":"Runway rolls out new AI video model that beats Google, OpenAI in key benchmark"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"https://www.cnbc.com/2025/12/01/runway-gen-4-5-video-model-google-open-ai.html"}},"_tags":["comment","author_xnx","story_46107738"],"author":"xnx","comment_text":"This is press release spam. Better to link to the announcement: https://runwayml.com/research/introducing-runway-gen-4.5","created_at":"2025-12-01T18:39:50Z","created_at_i":1764614390,"objectID":"46111187","parent_id":46107738,"story_id":46107738,"story_title":"Runway rolls out new AI video model that beats Google, OpenAI in key benchmark","story_url":"https://www.cnbc.com/2025/12/01/runway-gen-4-5-video-model-google-open-ai.html","updated_at":"2026-03-05T23:04:50Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bastawhiz"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"Have you tried adding Runway Gen-4 Turbo to Pro++?"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"I used ChatGPT to learn programming from zero and built a video generation SaaS"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.vidmakerpro.com/"}},"_tags":["comment","author_bastawhiz","story_44279325"],"author":"bastawhiz","comment_text":"Have you tried adding Runway Gen-4 Turbo to Pro++?","created_at":"2025-06-15T02:10:17Z","created_at_i":1749953417,"objectID":"44280084","parent_id":44279325,"story_id":44279325,"story_title":"I used ChatGPT to learn programming from zero and built a video generation SaaS","story_url":"https://www.vidmakerpro.com/","updated_at":"2025-06-15T05:56:14Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"isaacfrond"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"With new Gen-4 model, Runway has achieved consistency in AI videos"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"https://arstechnica.com/ai/2025/03/with-new-gen-4-model-runway-claims-to-have-finally-achieved-consistency-in-ai-videos/"}},"_tags":["story","author_isaacfrond","story_43544632"],"author":"isaacfrond","created_at":"2025-04-01T09:26:17Z","created_at_i":1743499577,"num_comments":0,"objectID":"43544632","points":2,"story_id":43544632,"title":"With new Gen-4 model, Runway has achieved consistency in AI videos","updated_at":"2025-04-01T10:15:34Z","url":"https://arstechnica.com/ai/2025/03/with-new-gen-4-model-runway-claims-to-have-finally-achieved-consistency-in-ai-videos/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"xnx"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"Runway is pretty far down on this leaderboard (Google Veo 2 is on top): https://huggingface.co/spaces/ArtificialAnalysis/Video-Gener...
Hopefully Gen-4 can improve."},"story_title":{"fullyHighlighted":true,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"Runway Gen-4"},"story_url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"https://runwayml.com/research/introducing-runway-gen-4"}},"_tags":["comment","author_xnx","story_43536085"],"author":"xnx","children":[43537323,43541083],"comment_text":"Runway is pretty far down on this leaderboard (Google Veo 2 is on top): https://huggingface.co/spaces/ArtificialAnalysis/Video-Gener...
Hopefully Gen-4 can improve.","created_at":"2025-03-31T16:06:01Z","created_at_i":1743437161,"objectID":"43536580","parent_id":43536085,"story_id":43536085,"story_title":"Runway Gen-4","story_url":"https://runwayml.com/research/introducing-runway-gen-4","updated_at":"2025-04-02T23:23:57Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vlovich123"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["runway","gen","4"],"value":"> despite being fairly safe, is still a risk compared to fusion
I recommend you read up on the Gen IV reactor designs. They\u2019re totally safe - meltdowns are impossible because of the way the reactor is built. If anything catastrophic happens, the reaction stops and can never get to a runaway reaction (physically impossible). Look up Gen 4 reactors. Those will be available before fusion even gets off the ground (and I\u2019ll note that fusion has 0 reactors built so who knows what kind of safety issues actually come up when engineering theory hits the road).
Even Gen III reactors are fine to put up everywhere (20x margin over Gen II) and Gen III+ reactors continue with the theme of adding passive safety measures that would prevent accidents like Fukushima and Chernobyl. Critics who rate any possibility of accident as unacceptable will never be pleased but that\u2019s not a reasonable position to take because nuclear energy isn\u2019t built in a vacuum and global warming and existing coal power poses a significantly higher threat and renewables and batteries simply can\u2019t scale no matter how hard we believe.
Fukushima and Chernobyl were Gen II designs which do have a cost advantage and EVEN WITH THOSE ACCIDENTS those designs are safer than existing coal and LNG power plants we are fine with having all over the place (nuclear is slightly safer than wind). Even Gen II designs built today are a fair bit safer than Chernobyl and Fukushima. Fukushima also ignored many and repeated safety warnings from internal and external reports although critics will generally point to this as a general criticism against all reactors (even though Fukushima still failed comparatively harmlessly all things considered).
Even with all of that, the death rate per kWH generated is drastically safer than coal and on par with wind and solar. Also construction costs tend to go down when the regulatory environment doesn\u2019t inhibit building reactors due to political fears that aren\u2019t grounded in the actual engineering.
I\u2019ll also note that China is building many many nuclear reactors and Russia is also following suit. So from a competition/national security perspective, China and Russia both have access to significantly more clean energy and more energy independence than we do.
Look. I understand there are problems with fission reactors. They remain the only feasible way to generate nuclear power in the next 60-100 years at scale. Yes there are downsides and risks. However there\u2019s one big upside vs fusion: it exists. It\u2019s possible to build these plants now without physics and engineering breakthroughs we haven\u2019t made yet. The advantages of fusion are safety, nuclear waste management, theoretical proliferation concerns. There\u2019s no reason to believe construction costs will be significantly lower. Even if they are, we\u2019re not even close to the first real commercial power plant even with this achievement as impressive as it is from a progress perspective."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"US Department of Energy: Fusion Ignition Achieved"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.energy.gov/articles/doe-national-laboratory-makes-history-achieving-fusion-ignition"}},"_tags":["comment","author_vlovich123","story_33971377"],"author":"vlovich123","children":[33994968],"comment_text":"> despite being fairly safe, is still a risk compared to fusion
I recommend you read up on the Gen IV reactor designs. They\u2019re totally safe - meltdowns are impossible because of the way the reactor is built. If anything catastrophic happens, the reaction stops and can never get to a runaway reaction (physically impossible). Look up Gen 4 reactors. Those will be available before fusion even gets off the ground (and I\u2019ll note that fusion has 0 reactors built so who knows what kind of safety issues actually come up when engineering theory hits the road).
Even Gen III reactors are fine to put up everywhere (20x margin over Gen II) and Gen III+ reactors continue with the theme of adding passive safety measures that would prevent accidents like Fukushima and Chernobyl. Critics who rate any possibility of accident as unacceptable will never be pleased but that\u2019s not a reasonable position to take because nuclear energy isn\u2019t built in a vacuum and global warming and existing coal power poses a significantly higher threat and renewables and batteries simply can\u2019t scale no matter how hard we believe.
Fukushima and Chernobyl were Gen II designs which do have a cost advantage and EVEN WITH THOSE ACCIDENTS those designs are safer than existing coal and LNG power plants we are fine with having all over the place (nuclear is slightly safer than wind). Even Gen II designs built today are a fair bit safer than Chernobyl and Fukushima. Fukushima also ignored many and repeated safety warnings from internal and external reports although critics will generally point to this as a general criticism against all reactors (even though Fukushima still failed comparatively harmlessly all things considered).
Even with all of that, the death rate per kWH generated is drastically safer than coal and on par with wind and solar. Also construction costs tend to go down when the regulatory environment doesn\u2019t inhibit building reactors due to political fears that aren\u2019t grounded in the actual engineering.
I\u2019ll also note that China is building many many nuclear reactors and Russia is also following suit. So from a competition/national security perspective, China and Russia both have access to significantly more clean energy and more energy independence than we do.
Look. I understand there are problems with fission reactors. They remain the only feasible way to generate nuclear power in the next 60-100 years at scale. Yes there are downsides and risks. However there\u2019s one big upside vs fusion: it exists. It\u2019s possible to build these plants now without physics and engineering breakthroughs we haven\u2019t made yet. The advantages of fusion are safety, nuclear waste management, theoretical proliferation concerns. There\u2019s no reason to believe construction costs will be significantly lower. Even if they are, we\u2019re not even close to the first real commercial power plant even with this achievement as impressive as it is from a progress perspective.","created_at":"2022-12-14T13:04:11Z","created_at_i":1671023051,"objectID":"33982888","parent_id":33980659,"story_id":33971377,"story_title":"US Department of Energy: Fusion Ignition Achieved","story_url":"https://www.energy.gov/articles/doe-national-laboratory-makes-history-achieving-fusion-ignition","updated_at":"2024-09-20T12:53:32Z"}],"hitsPerPage":20,"nbHits":12,"nbPages":1,"page":0,"params":"query=runway+gen-4&advancedSyntax=true&analyticsTags=backend","processingTimeMS":34,"processingTimingsMS":{"_request":{"roundTrip":20},"afterFetch":{"format":{"highlighting":1,"total":1},"merge":{"mergeLoop":{"prepareNextHit":2,"total":2},"total":2},"total":2},"fetch":{"query":6,"scanning":24,"total":31},"total":34},"query":"runway gen-4","serverTimeMS":36}