Cloudwalk Congrong 2
Cloudwalk · Undisclosed (CV lineage; vision-tower led)
The only CV “little dragon” still shipping multimodal
The pivot product of one of the CV “little dragons”: perception 91 is its old trade, everything else is support. Composite 76.1, ranked 32nd — the lowest here — yet it remains steady on recognition-rather-than-understanding tasks like faces, receipts and ID documents. It is the only model of 32 whose perception exceeds its own composite by more than 15 points.
Last updated Oct 18, 2026Editorial scoring
Bottom line
Eight-dimension breakdown
editorial mapping · 0–100- Cloudwalk Congrong 276.1
Weights: perception 20 / documents 18 / charts 14 / video 14 / agents 14 / generation 6 / Chinese 6 / ecosystem 8. Definitions and sources on the methodology page.
Public benchmarks
as reported, not converted| Benchmark / metric | Value | Note |
|---|---|---|
| 人像识别误识率 | 1e-6 | Vendor figure |
| 输出单价 | ¥7 | Industry pricing |
Public scores come from different versions and prompts and are not interchangeable with our composite; they are one evidence input.
Scenario mapping
five stress tests, one scale| Scenario | Score | Observation |
|---|---|---|
| Chinese receipt OCR | 80 · Steady recognition, average layout | |
| Chart data extraction | 68 · Not a target scenario | |
| 30-min video QA | 62 · No video modality | |
| Mobile GUI control | 64 · Not trained for GUI | |
| Fine-grained recognition | 91 · Same as our perception dimension |
Strengths and limitations
Strengths
- Perception 91: face, ID and receipt recognition stability remains near the front.
- Deep industry channels: bank branches, public security and campus delivery networks are ready-made.
- Moderate pricing: ¥7 industry pricing, below most score peers.
Limitations
- Composite 76.1 is the lowest here: weak across the board except perception.
- No video modality — long-horizon tasks are absent outright.
- Ecosystem 68: SDK and docs target industry customers; poor developer experience.
Price and real cost
List price
¥7/M tokens
≈ ¥7 / 1M output tokens (industry pricing). For 1,000 images at 1,024px, ~1,100 image tokens each and 300 output tokens, the monthly cost is about ¥4.2 (assuming input at 25% of output price).
Use your volumeAbout Cloudwalk Congrong 2
Is Cloudwalk Congrong 2 open source?
Licence: Closed commercial. No public weights; access is via vendor API or an on-prem licence.
How much does Cloudwalk Congrong 2 cost?
≈ ¥7 / 1M output tokens (industry pricing). It ranks #32 with a composite of 76.1. Real bills depend on resolution, tiling, caching and volume discounts.
What is Cloudwalk Congrong 2 best for?
ID verification and receipt recognition; Bank-branch and campus recognition scenarios; Cost baseline for recognition tasks。
Why does Cloudwalk Congrong 2 score 76.1?
The composite weights eight dimensions: perception 20%, documents 18%, charts 14%, video 14%, agents 14%, generation 6%, Chinese 6%, ecosystem 8%. Its strongest dimension is Visual perception & grounding; its weakest is Generation & editing.
Personal ranking · Not an official score
Every ranking, score and comparison on this page is an independent personal ranking, not an official score, and represents no vendor, institution or regulator. Scores are a subjective mapping of public evidence under our own weights — change the weights and the order changes. Disagree, or spotted a factual error? Contact the site owner: jizhi2024@foxmail.com — we check each report and log corrections in the changelog.