We're excited to share that we've already tested GPT-4 Vision 128K and wanted to share it with you.
Since access to the GPT-4V (Vision) model is currently very limited on OpenAI's side, we'll be selecting some users who actively use GPT-4 on our platform to include in the testing group. The GPT-4-Vision 128K model will cost 3 times less than GPT-4, while running 2 times faster! It can take images for analysis and process a huge amount of context.
Starting today, gptunnel users get access to the GPT-4V (Vision) model. We currently have a daily limit in place, but as soon as the model leaves preview, we'll immediately make access to GPT-4 Vision unlimited! 😎
What can GPT-4V do?
The main new feature is the ability to analyze images with GPT-4V. We can pass one or several images and ask a question about them, for example:
The second important feature of GPT-4V-128K is its context window, which is 16 times larger than the regular GPT-4. This means the model can process far more text than previous models.
We've also already noted that the model has become 3 times cheaper — its price with us is $0.03 per 1000 context tokens.
All in all, we now have a model that runs faster, processes more text, and costs less.
When will the model be available?
Within a few weeks, we'll lift the limit, but in the meantime, to get access to the preview version of the model, you can sign up and start using the GPT-4V model right now!
How to pay for GPT-4V (Vision)?
Once you get access to the coveted GPT-4-Turbo-Vision-128K model (its full name from OpenAI), you'll be able to simply top up your balance for any amount using local payment methods and start using it!
Just a reminder that we also work with business clients on a b2b basis, and they too have a chance to get access to the model today! To learn more about corporate cooperation with us, check out the information on the business page.