Choose Pika H3 Flex when…
Choose Pika H3 Flex if you want to self-host open weights or pay per output second through Pika's API, and your shots fit inside 15 seconds with audio you need to remix stem by stem.
Video model alternative
Pika H3 Flex Alternative: What You Can Generate on Imgo Today
Pika H3 Flex is Pika's post-trained MiniMax H3 video model, launched October 29, 2026 with open weights, exportable audio stems and a per-second API. It is not available on Imgo. Compare its published rates and limits with Seedance 2.0, which you can run on Imgo today.
External reference — not available on ImgoLast checked: 2026-11-03
The launch window closed on November 5, 2026. The halved rates and the daily free generations ended with it; the figures on this page are Pika's regular list, last checked on 2026-11-03 — re-read Pika's product page before you budget.
Reviewed against linked first-party pages. Plan limits and model availability can change.
Compare the working requirements
Choose Imgo when…
Choose Imgo when you need to produce today in a hosted workspace: Seedance 2.0 renders up to 15 seconds with native audio from text, a start frame, a start and end frame, or references, next to image generation, editing and a shared asset library. Pika H3 Flex itself is not available on Imgo.
Compare the working requirements
The table describes documented capabilities and checks to make, without assigning unsupported scores or assuming a plan includes every tool.
| Decision factor | What to verify |
|---|---|
| Best suited to | Creators and developers weighing Pika's open-weight, per-second video model against a hosted workspace they can already use — especially anyone budgeting a project inside the November 5 launch window. |
| Image and video | Pika H3 Flex is a video model served through Pika's API and, since November 2, 2026, Pika's creation product — or self-hosted from the open weights. Imgo hosts several video models, plus image generation and editing for the frames that feed them. |
| References and editing | H3 Flex takes up to 12 reference assets — at most 9 images, 3 videos and 3 audio files, with video and audio references capped at 15 seconds — and edits an uploaded clip by prompt while preserving the performance and timing. Seedance 2.0 on Imgo takes image, video and audio references, and Imgo's video editor revises a finished clip by instruction. |
| Character and brand consistency | Pika's speed and leaderboard claims are its own until you test them: the roughly 7-second draft figure is transformer compute in Pika's controlled benchmark, not end-to-end API latency. Run the same character and product references through H3 Flex and Seedance 2.0, and compare the face and wardrobe in the last second against the first. |
| API, batch, and team | Pika serves H3 Flex through its API with per-second billing and added it to the Pika creation product on November 2, 2026; self-hosting means running the open weights yourself. Imgo exposes batch, team and API surfaces around the models it hosts. |
| Price and free use | Pika bills H3 Flex per output second, tiered by resolution from native 480p up to upscaled 4K, with a higher schedule when a video reference adds its input seconds to the bill; audio stems add a per-second surcharge and the prompt rewriter a per-request fee. Through November 5, 2026, launch pricing halves the per-second rates and Pika API Club members get 15 free generations a day — read the live figures on Pika's model page. H3 Flex is not priced on Imgo, where every run quotes its credits live before you submit. |
Where Pika H3 Flex is strong
Pika lists 5–15 second clips at 24 fps in eleven aspect ratios from 21:9 to 9:21, prompts up to 8,000 characters, native stereo audio, and five exportable stems — dialogue, score, SFX, foley and ambience. A generation takes up to 12 reference assets (as many as 9 images, 3 videos and 3 audio files), drafts at 480p or 768p, and finishes through an in-job upsampler at up to 4K. The weights are public on Hugging Face and GitHub: generator adapters and inference code covering text-to-video, first- and last-frame, multimodal reference-to-video and editing.
Source: Pika: H3 Flex model page
A reproducible test for your team
Suggested test conditions; no independent benchmark result is claimed.
Prompt
A 12-second 16:9 take with audio: a chef in a white coat plates a dessert behind a marble counter while the camera dollies in from a wide shot to a medium close-up; she says "Finish it with the berry glaze," the whisk taps the bowl twice, and kitchen ambience sits under the dialogue.
Input and controls
Use the same chef reference image and 16:9 framing on both paths and generate three takes each at 12 seconds: H3 Flex through Pika's API and Seedance 2.0 on Imgo. Keep every take; export H3 Flex's dialogue stem and check the line against the video.
Record these outcomes
- Camera dolly completed without a cut
- Face and coat identical in the last second and the first
- Dialogue line intelligible and in sync
- Cost per accepted take
Check live pricing before you decide
Not priced on Imgo
Plans, free usage, API access, and model limits change. Compare the current pages and run the same test batch on both services.
Frequently asked questions
- Can I use Pika H3 Flex on Imgo?
- No. Pika H3 Flex is not integrated into Imgo and has no Imgo credit price; this page is an external reference, last checked on 2026-11-03. You can run it through Pika's API, in Pika's product, or self-hosted from the open weights. Seedance 2.0 is the video model to run on Imgo meanwhile.
- What does Pika H3 Flex cost?
- Pika bills per output second: $0.02 at 480p, $0.04 at 768p, $0.07 at 1080p, $0.08 at 2K and $0.16 at 4K; reference-to-video from a video input bills input and output seconds at $0.06 to $0.20 across the same resolutions. Audio stems add $0.001 per generated second and the prompt rewriter $0.02 per request. Through November 5, 2026, the per-second rates are halved and API Club members get 15 free generations a day.
- How fast is Pika H3 Flex really?
- Pika's published figure is about 7 seconds of transformer compute for a 15-second clip with synchronized audio at 768p, from a controlled benchmark rather than end-to-end API latency. Time it on your own prompts before scheduling work around it.
- Is Pika H3 Flex open source?
- The generator adapters and inference code are public on Hugging Face and GitHub — enough to run text-to-video, first- and last-frame, reference-to-video and editing yourself. Pika's hosted finishing features, the up-to-4K upsampler and audio-stem export, are not part of the weights.