whatmodel.sh / video / Kling 2.5 Turbo Pro
Kling 2.5 Turbo Pro
↗ share
+ contribute
Kling 2.5 Turbo Pro is a closed-weights video model from Kuaishou, released September 2025.
Model information sources
$ how it compares
The video models ranked either side of Kling 2.5 Turbo Pro, out of 139 with a score, then more from Kuaishou.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 38 |  Cosmos3-Super | 58±4 | local | ||
| 39 |  Seedance 1.5 Pro | 58±2 | api $0.052/sec | ||
| 40 | Kling 2.5 Turbo Pro | 57±2 | api $0.070/sec | ||
| 41 |  Runway Aleph | 56±5 | comparison only | ||
| 42 |  Wan 2.2 I2V A14B | 56±2 | api $0.080/sec +local | ||
| more from Kuaishou | |||||
| 18 | Kling 3.0 Pro | 69±2 | api $0.11/sec | ||
$ what it's good at
text to video▆▄▃▄▆▅█▃▂▂▂▄ #20 of 68 * · 56% of its score · 2 tests
Beats Kling 2.6 Pro, Cosmos3-Super here, though they rank higher overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| LMArena video | 1219.8 ✎ | ▂▂▃▆█▆·▆▅▁▂▇ beats 19/41 | 54% |
| HumanScore (T2V) | 90.8% ✎ | ▂··▂··█▄▂··▆ beats 9/11 | 1.6% |
image to video▁▂·▂▅█▅▄▅▆▃▄ #29 of 50 · 43% of its score · 3 tests
Beats Seedance 1.5 Pro here, though they rank higher overall. Trails Cosmos3-Nano, Seedance 1.0 Pro here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Arena image-to-video | 1275.6 ✎ | ▁▁▁▁▄▇▃▆▃▂▄█ beats 17/40 | 36% |
| VWG-Bench think-with-video reasoning (overall) lab | 59.4% ✎ | █··▄······▄█ beats 2/5 | 5.4% |
| V-ReasonBench (average pass@5) | 11.3% ✎ | █·····▄···▄▄ beats 1/4 | 1.8% |
$ where to run it
Through an API
api · fal.ai · $0.070/sec · fal.ai ↗ · price sourceUSD per second of generated video on fal.ai. Product: fal Kling 2.5 Turbo Pro text-to-video endpoint. API model: fal-ai/kling-video/v2.5-turbo/pro/text-to-video. Price basis: fal prices a 5-s video at $0.35 and each additional second at $0.07, i.e. $0.07 per output second. Kling's own developer pricing page does not list a per-model price for V2.5 Turbo. Currency: USD. fal.ai model page, observed 2026-09-25.
$ how the score was worked out · LMArena leaderboard dataset 90%, From Evaluation to Enhancemen… 5%, V-ReasonBench: Toward Unified… 2% and 1 more
Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works
| source | results | share of the score |
|---|---|---|
 LMArena leaderboard dataset | 2 | 90% |
 From Evaluation to Enhancement: Benchmarking and Improving Think-with-Video Reasoning for Video Generative Models (VWG-Bench) · lab-reported | 1 | 5% |
| V-ReasonBench: Toward Unified Reasoning Benchmark Suite for Video Generation Models | 1 | 2% |
| HumanScore: Benchmarking Human Motions in Generated Videos | 1 | 2% |
| prior · pull toward the middle | 1% | |
| weighted average | = 57.1 |
$ improve this page
Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.
data updated · sources and licences · json