PikaVideo model
Pika 2.2
In image to video, Pika 2.2 is in place 21 of 21, with text only in place 23. A 5 s clip costs about $0.20.
997Arena ratinghigher is better
1,009Arena ratinghigher is better
$0.20$0.04 per secondlower is better
n/alongest single cliphigher is better
17out of 100same as last month
Summary
At $0.04 per second of video it is cheap; a typical video model costs $0.12.
What is distinctive about Pika 2.2 is Pikaframes. The image-to-video feature fixes a first and a last image as endpoints, while text-to-video and standard image-to-video run at 720p or 1080p.
Other 2.2 generations, including Pikascenes, last 5 or 10 seconds. Pikaframes reaches up to 25 seconds and uses the two uploaded still images as the beginning and end of the clip.
Pika 2.2 benchmarks
Image to video arena
How people rate generated videos in blind comparisons, Elo points · higher is better
20 of 21 models
as of September 22, 2026 · Source: LMArena
Text to video arena
How people rate generated videos in blind comparisons, Elo points · higher is better
20 of 23 models
as of September 22, 2026 · Source: LMArena
Show all 28 individual results from 8 sources
Page 1 of 3
Published benchmarks
LMArena Text to Video
Elo ratings summarize pairwise comparisons within this leaderboard. They are not percentages. The scale does not express absolute quality or linear gaps between models.
higher is betterAs of November 5, 2026Scale: 0 Elo - 1,516 Elo
Methodology and context
- Metric
- elo-rating
- Scale
- No upper bound is documented. The chart spans the recorded values and zero.
- Participants
- 23
- Sample
- 5,763
LMArena Image to Video
Elo ratings summarize pairwise comparisons within this leaderboard. They are not percentages. The scale does not express absolute quality or linear gaps between models.
higher is betterAs of November 5, 2026Scale: 0 Elo - 1,495 Elo
Methodology and context
- Metric
- elo-rating
- Scale
- No upper bound is documented. The chart spans the recorded values and zero.
- Participants
- 21
- Sample
- 8,635
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.61
- Score
- 0.61
- Metric
- Average score (0-1), physics
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- physics
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.73
- Score
- 0.73
- Metric
- Average score (0-1), nature
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- nature
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.6
- Score
- 0.6
- Metric
- Average score (0-1), activity
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- activity
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.57
- Score
- 0.57
- Metric
- Average score (0-1), culture
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- culture
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.56
- Score
- 0.56
- Metric
- Average score (0-1), causality
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- causality
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.56
- Score
- 0.56
- Metric
- Average score (0-1), object
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- object
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VWorldBench (Show measurement, test conditions, and source)
- Source value
- 0.6
- Score
- 0.6
- Metric
- Average score (0-1), overall
- Unit
- ratio
- Benchmark version
- arXiv:2507.18107v1, Table 2
- Category
- overall
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VWorldBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), T2VWorldBench Table 2; the paper describes six domains with 200 prompts each, typically the lowest available resolution (usually 720p), 16:9 aspect ratio, and about five seconds. The published model row uses human and automatic evaluation (two raters; LLaVA 1.6 34B).
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
SeqBench (Show measurement, test conditions, and source)
- Source value
- 0.69
- Score
- 0.69
- Metric
- DTG average score
- Unit
- ratio
- Benchmark version
- arXiv:2510.13042v1, Table I
- Category
- temporal-detail-generation
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- SeqBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact Pika 2.2 row in the paper's text-to-video cohort; the benchmark has 320 prompts and reports 2,560 videos across eight models. DTG and narrative-coherence results are separate aggregate metrics. The paper reports five human evaluators per video for its human checks, but does not publish a Pika-specific denominator for either result.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
SeqBench (Show measurement, test conditions, and source)
- Source value
- 0.228
- Score
- 0.228
- Metric
- Narrative coherence overall score
- Unit
- ratio
- Benchmark version
- arXiv:2510.13042v1, Table II
- Category
- narrative-coherence
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- SeqBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Exact Pika 2.2 row in the paper's text-to-video cohort; the benchmark has 320 prompts and reports 2,560 videos across eight models. DTG and narrative-coherence results are separate aggregate metrics. The paper reports five human evaluators per video for its human checks, but does not publish a Pika-specific denominator for either result.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
T2VPhysBench (Show measurement, test conditions, and source)
- Source value
- 0.38
- Score
- 0.38
- Metric
- Main physical-consistency score (0-1), Newton's laws
- Unit
- ratio
- Benchmark version
- arXiv:2505.00337v1, Table 2 (main setting)
- Category
- newton's-laws
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- T2VPhysBench
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Independent evaluation. Text-to-video (T2V), exact Pika 2.2 row in T2VPhysBench. The main table evaluates physical consistency; the separate counterfactual table uses prompts that deliberately request physically impossible events. Output controls and a per-model N are not reported in the score tables.
- Context
- This is a task-specific paper result, not a universal video-quality score. Do not compare its numeric value directly with other benchmark scales.
Recorded measurement details
LMArena Text to Video: LMArena Text to Video (Show measurement, test conditions, and source)
- Source value
- 1009
- Score
- 1009
- Metric
- elo-rating
- Unit
- Elo
- Task
- lmarena-text-to-video
- Task label
- LMArena Text to Video
- Direction
- Higher is better
- Source rank
- 47
- Sample
- 5763
- Lower confidence bound
- 994
- Upper confidence bound
- 1024
- Retrieved at
- 2026-11-05
LMArena Image to Video: LMArena Image to Video (Show measurement, test conditions, and source)
- Source value
- 997
- Score
- 997
- Metric
- elo-rating
- Unit
- Elo
- Task
- lmarena-image-to-video
- Task label
- LMArena Image to Video
- Direction
- Higher is better
- Source rank
- 48
- Sample
- 8635
- Lower confidence bound
- 983
- Upper confidence bound
- 1011
- Retrieved at
- 2026-11-05
Page 1 of 3
Video sample from our test
This video exists as a documented run for the exact model version. Playback is muted.
Motion and physical consistency
This sample tests Anatomy, object consistency, camera tracking, rain, and reflections.
Prompt and settings
Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.
Generated through fal.ai with model fal-ai/pika/v2.2/text-to-video. Archived settings were 16:9, one request, first output, requested 5 s, seed provider random, request {"prompt":"Generate exactly one continuous five-second 16:9 cinematic video. A fictional bicycle courier in a mustard-yellow raincoat rides a red city bicycle from left to right across a rain-wet Hamburg street at blue hour for the entire clip. The camera tracks her smoothly at waist height. Both hands remain visible on the handlebars, the bicycle geometry stays physically correct, and a transparent umbrella fixed to the bicycle moves naturally in the wind. Wet-street reflections and raindrops follow realistic physics. No cuts, no slow motion, no text, no logos, no subtitles, and no audio.","aspect_ratio":"16:9","resolution":"720p","duration":5}, configured price estimate $0.20.
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
0 to 100, higher is more popular
20 of 98 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 5, 2026
Pika 2.2Kling 3.0 ProHailuo 2.3Dreamina Seedance 2.5 720pSora 2 Pro
Pika 2.2 compared
Compare with another model
The most popular comparisons
- 1Gemini Omni Flash vs. Pika 2.2
- 2FLUX.3 Video vs. Pika 2.2
- 3Dreamina Seedance 2.5 720p vs. Pika 2.2
- 4Dreamina Seedance 2.0 720p vs. Pika 2.2
- 5MiniMax H3 vs. Pika 2.2
- 6Sora 2 Pro vs. Pika 2.2
- 7Sora 2 vs. Pika 2.2
- 8Veo 3.1 vs. Pika 2.2
- 9Veo 3.1 Fast vs. Pika 2.2
Show all 48 comparisons
- 10Grok Imagine Video 1.5 vs. Pika 2.2
- 11Grok Imagine Video vs. Pika 2.2
- 12Wan 2.7 vs. Pika 2.2
- 13Wan 2.6 vs. Pika 2.2
- 14Runway Gen-4.5 vs. Pika 2.2
- 15Kling 3.0 Pro vs. Pika 2.2
- 16Kling 2.6 Pro vs. Pika 2.2
- 17Vidu Q3 Pro vs. Pika 2.2
- 18PixVerse V5.6 vs. Pika 2.2
- 19Hailuo 2.3 vs. Pika 2.2
- 20LTX-2 19B vs. Pika 2.2
- 21Wan 2.2 A14B vs. Pika 2.2
- 22HunyuanVideo 1.5 vs. Pika 2.2
- 23Pika 2.2 vs. Gemini Omni 1.1 Flash
- 24Pika 2.2 vs. Muse Video
- 25Pika 2.2 vs. Wan 3.0
- 26Pika 2.2 vs. Wan 3.0 Prime
- 27Pika 2.2 vs. HappyHorse 1.0
- 28Pika 2.2 vs. HappyHorse 1.1
- 29Pika 2.2 vs. PixVerse V6
- 30Pika 2.2 vs. PixVerse C1
- 31Pika 2.2 vs. Seedance 2.0 Fast
- 32Pika 2.2 vs. Seedance 2.0 Mini
- 33Pika 2.2 vs. Vidu Q3 Turbo
- 34Pika 2.2 vs. Vidu Q2
- 35Pika 2.2 vs. Vidu Q1
- 36Pika 2.2 vs. Kling 4.0
- 37Pika 2.2 vs. Kling 3.0 Standard API
- 38Pika 2.2 vs. Kling Video O1
- 39Pika 2.2 vs. Hailuo 2.3 Fast
- 40Pika 2.2 vs. Wan 2.5 Preview
- 41Pika 2.2 vs. Wan 2.6 I2V Flash
- 42Pika 2.2 vs. Wan 2.2 TI2V 5B
- 43Pika 2.2 vs. Wan 2.2 I2V A14B
- 44Pika 2.2 vs. HunyuanVideo 1.0 T2V
- 45Pika 2.2 vs. HunyuanVideo I2V 1.0
- 46Pika 2.2 vs. Ray 3.2
- 47Pika 2.2 vs. Grok Imagine Video 1.5 Lite
- 48Pika 2.2 vs. LTX-2.5
All specifications
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Frequently asked questions about Pika 2.2
How good is Pika 2.2?
Rank 23 of 23 in the text-to-video arena rating. With a start image: rank 21 of 21.
LMArena, as of Sep 22, 2026
How much does a video with Pika 2.2 cost?
About $0.20 for a 5 s clip, $0.04 per second of video. That is cheap.
Sources and data date
Show all 21 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| LMArena | 0 | 0 | 2 |
Without vendor claims and historical or disputed results; those stay labelled in the single values. 2 measurements are not yet assigned to a test series.
| Type | Evidence and data date |
|---|---|
| Research status | Research date October 16, 2026. 16 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not found in the inspected sources | wiro-video-models-exact-row Note The inspected Wiro leaderboard has no exact model-version route. Nearby variants were not transferred. |
| Not found in the inspected sources | arena-video-edit-exact-row Note The inspected Video Edit table has no exact model-version row. T2V and I2V values were not reused. |
| Not found in the inspected sources | vabench-exact-model-row Note VABench has no exact model row. Sora 2 was not transferred to Sora 2 Pro or other models. |
| Not found in the inspected sources | artificial-analysis-t2v-exact-row Note No exact model version or resolution was found in the inspected Artificial Analysis table. |
| Not found in the inspected sources | artificial-analysis-i2v-exact-row Note No exact model version or resolution was found in the inspected Artificial Analysis table. |
| Unresolved | vbench-current-public-leaderboard Note The current VBench leaderboard is embedded in a dynamic Hugging Face iframe and was not retrievable; current scores remain unresolved. |
| Not found in the inspected sources | vidbenchmark-exact-row Note No exact model id is present in the 2026-08-04 dataset; similarly named variants were not transferred. |
| Additional source | Pika · Pika official platform · Editorial description reviewed October 2, 2026 |
| Additional source | Video pricing |
| Additional source | fal.ai |
| Additional source | LMArena (retrieved November 5, 2026)12 |
| Additional source | T2VWorldBench (retrieved October 16, 2026) |
| Additional source | SeqBench (retrieved October 16, 2026) |
| Additional source | T2VPhysBench (retrieved October 16, 2026) |
| Additional source | SuperCLUE-I2V, 2025 H1 report, printed page 49 (retrieved October 17, 2026) |
| Additional source | Arena Video Image-to-Video (retrieved September 21, 2026; October 16, 2026; October 28, 2026) |
| Additional source | Arena Video Text-to-Video (retrieved September 21, 2026; October 16, 2026; October 28, 2026) |
| Additional source | Model metadata (retrieved October 16, 2026)123 |
| Additional source | Pika FAQ · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |