how do I obtain keys for computer vision 2.0

[icistech]

I subscribed to computer vision to get access to the improved OCR in 2.0, but it looks like I can only purchase a 1.0 key. How do i get a computer vision 2.0 key (free or paid)?

---

Document Details

⚠ Do not edit this section. It is required for docs.microsoft.com ➟ GitHub issue linking.

• ID: 8b224939-86a1-e97c-dfe6-b466acf30a85
• Version Independent ID: d4c66f21-3f6e-b09a-35e5-84d8953b1e87
• Content: Get subscription keys for the Computer Vision API
• Content Source: articles/cognitive-services/Computer-vision/Vision-API-How-to-Topics/HowToSubscribe.md
• Service: cognitive-services
• GitHub Login: @KellyDF
• Microsoft Alias: kefre

[SaurabhSharma-MSFT]

@icistech Thanks for your feedback! We will investigate and update as appropriate.

[YutongTie-MSFT]

@icistech Thanks for the feedback! I have assigned the issue to the content author to investigate further and update the document as appropriate.

@cjgronlund Hi, do you have any idea about the 2.0 key? Thanks a lot!

[arightious]

I need to know this as well. In my console there is no option to set the version, endpoint or region.

[cjgronlund]

Please assign to @noellelacharite

in-progress

[noellelacharite]

I am working with the service team to investigate.

in-progress

[sheheryarnaveed]

Yeah I am facing the same problem. The api used on microsoft computer vision api website gives better result for OCR as compared to the v1.0 version api given on cognitive services. I believe the difference of v1.0 and v2.0 is the reason behind that. May I know how long is it going to take for getting the v2.0 version to get public?

[gpkc]

+1, @noellelacharite would you possibly have updates on this? :)

[KellyDF]

Hi all, All keys that you get from the Azure portal will work for v2.0. Call the API with v2.0 at the end instead of v1.0. For instance https://southcentralus.api.cognitive.microsoft.com/vision/v2.0.

We are aware of the UX bug and fixing it soon.

Thanks, Kelly

[crs2m5smith]

I have been trying for days to get this to work. I want to use version 2.0 Api and every time I get access denied. Some how can you show me a new article or guide on how exactly to do this. My v1.0 simply does not work for the v2.0 West Central US endpoint for Vision

[crs2m5smith]

Update ? @noellelacharite

[crs2m5smith]

Were you able to get a resolution on your issue too? @sheheryarnaveed

[gpkc]

@crs2m5smith I was having a similar issue. Turned out I was setting the wrong region. Did you check if you're using the same region for both v1.0 and v2.0?

[crs2m5smith]

@gpkc Ya pretty sure it is westcentralus. That is what the portal says. How do you check v2.0 though? The portal only show v1.0

[crs2m5smith]

@gpkc Ar you using the free version or paid version btw. I think I saw on the free version it showed a v2.0 endpoint but other than that , I have not been able to find anything on v2.0 api keys

[easyj2j]

@crs2m5smith Note that if you're using the paid version, the SO tier doesn't support v2.0

[crs2m5smith]

Maybe I am missing something... So this is the version I am trying to use: https://westcentralus.dev.cognitive.microsoft.com/docs/services/5adf991815e1060e6355ad44/operations/587f2c6a154055056008f200/console

After entering my key and example image I get a 202 response and when i go to the operation location in response and get: { "statusCode": 401, "message": "Access denied due to missing subscription key. Make sure to include subscription key when making requests to an API." }

The endpoint url is: https://westcentralus.api.cognitive.microsoft.com/vision/v2.0/recognizeText?mode=Printed

My service is in deed on pricing tier S1 and west central region

[easyj2j]

@crs2m5smith Just double-checking here but are you including the subs key in the second request?

[crs2m5smith]

@easyj2j I must be missing something. What second request? Do you have a link or example. This is my first request essentially HttpResponseMessage response; // Request body byte[] byteData = Convert.FromBase64String(base64Vin); using (var content = new ByteArrayContent(byteData)) { content.Headers.ContentType = new MediaTypeHeaderValue("application/octet-stream"); response = await client.PostAsync(uri, content); }

[easyj2j]

recognizeText requires 2 calls. The first returns the location of the results in the Operation-Location header. If you're only interested in printed text, you can use the ocr method, which requires only a single call. See https://docs.microsoft.com/en-us/azure/cognitive-services/Computer-vision/quickstarts/csharp-hand-text

[crs2m5smith]

@easyj2j you are the man! So I have been looking at the Printed word mode docs and example. and it does not mention or show the second call. The link you gave me was for the handwritten example and it showed the second call. Thanks. Everything works

[noellelacharite]

Thanks everyone, looks like this has been resolved. @SaurabhSharma-MSFT you can close this issue.

Thanks!

[ghost]

Does anyone know how i would do go about getting handwritten text using the microfost.projectoxford library instead?

[SaurabhSharma-MSFT]

@icistech We will now proceed to close this thread. If there are further questions regarding this matter, please reopen it and we will gladly continue the discussion.