Pika · video model

Pika H3 Flex

Pika H3 Flex is Pika's latest video-generation model in our dataset, released Oct 29, 2026. Open weights (MiniMax H3 Community License Agreement). Pika's open-weights video-with-audio model, post-trained from MiniMax H3 for six-step generation; released Oct 29, 2026, with local 480p/768p inference and up to 4K through Pika's API. Verified against vendor sources on Oct 31, 2026.

Reported Last verified · All models · JSON · Markdown

Share on X

Primary source: https://pika.art/blog/pika-h3-flex-research. We are the fast, independent index; the vendor’s own page is the source of record for these model facts.

Pika H3 Flex at a glance: takes text, image, video, audio in; returns video, audio.

  • ReleasedOct 29, 2026
  • LicenceOpen weights, restrictedMiniMax H3 Community License Agreement
InTextImageVideoAudioOutVideoAudio

Key facts

Vendor-reported facts are read from the linked vendor page on the date shown above; we did not run the model. Independently tested means we ran it; no fact on this page is. Editorial is our own classification, not a vendor claim. Where a linked page does not itself state a value, the row says so. How to read provenance.

Pika H3 Flex: key facts with sources
FactValueLabelSource
VendorPikaEditorialn/a
ReleasedOct 29, 2026 Date on Pika's launch post; the Hugging Face repository was created the same day.Vendor-reportedsource
Statusgenerally availableEditorialn/a
Inputtext, image, video, audio Text prompts plus up to 12 reference inputs, per the model card.Vendor-reportedsource
Outputvideo, audioVendor-reportedsource
LicenceMiniMax H3 Community License Agreement (open weights) The adapters are Model Derivatives of MiniMax H3 under its community licence; the Qwen encoder assets are Apache 2.0.Vendor-reportedsource

Where Pika H3 Flex sits in our data

Alternatives to Pika H3 Flex

Video models from other vendors that share at least one of our groups with Pika H3 Flex. Comparable on paper, from each model’s sourced facts; we have not run them.

  • Genie 4Google · Both in: Video · released Oct 27, 2026
  • Gemini Omni 1.1 FlashGoogle · Both in: Video · Input per 1M tokens $1.5; Output per second of 720p video (approx.) $0.1 (mixed) · released Aug 27, 2026

What the sources say

These are claims from the linked pages, not our measurements.

  • Pika says H3 Flex generates video with synchronized stereo audio in six sampling steps and takes up to 12 reference inputs. Vendor description; we did not run it.

Notes

  • Local inference covers native 480p and 768p; 1080p, 2K and 4K, audio stems and prompt enhancement are Pika API features.

What we tested

Nothing yet. We have not run Pika H3 Flex ourselves, so every fact on this page is Reported from the sources below. When we do test it, the scope will be stated here and in our test logs. See methodology.

Not verified, left out

We could not confirm these on a vendor page, so we do not state them.

  • Parameter count: the model card does not state one; the released files are adapters on top of MiniMax H3, so we do not report a size.
  • Benchmark tables: Pika's own; not reproduced here.
  • API pricing for the hosted 1080p, 2K and 4K tiers was not captured.

Pika H3 Flex FAQ

When was Pika H3 Flex released?

Pika released Pika H3 Flex on Oct 29, 2026. Source.

Is Pika H3 Flex open weights?

Yes. The weights are published under MiniMax H3 Community License Agreement. Source.

Has Agentic Weekly tested Pika H3 Flex?

Not yet. Every fact on this page is Reported: read from the linked vendor pages, not measured by us.

Sources

Compare and guides

We have no head-to-head page or guide that names Pika H3 Flex yet. These hubs cover the models we do compare.

    All model comparisons · All guides · Model releases by month

    The Weekly Verdict · free

    5 new-model use cases. Every Monday.

    A free email every Monday: the week's few big AI releases, then 5 short use cases for the new models, each with a prompt you can paste.

    Free. Double opt-in. One click to unsubscribe.