As of version 2026-05, Glob.AI OS optimized support for scanned and image-based PDFs with compatible models to include file input where available.
The platform metadata has been updated so that all supported models capable of processing PDF files through Chat API are now exposed with File support.
Glob.AI OS now correctly identify these models as accepting file inputs.
| Provider |
Models updated in Version 2026-05 |
| OpenAI |
GPT-5.5, GPT-5.4, GPT-5.4 Mini, GPT-5.4 Nano, GPT-5.2, GPT-5.1, GPT-5, GPT-5 Mini, GPT-5 Nano, GPT-4.1, GPT-4.1 Mini |
| AWS Bedrock |
Amazon Nova 2 Lite |
These models now support Text, Image, and File inputs.
The remaining supported providers (Vertex AI, Anthropic, and Azure AI Foundry) already exposed File support before this update, so no configuration changes were required.
| Provider |
Before |
After |
| OpenAI GPT-5 family |
Text, Image |
Text, Image, File |
| OpenAI GPT-4.1 family |
Text, Image |
Text, Image, File |
| AWS Bedrock Amazon Nova 2 Lite |
Text, Image |
Text, Image, File |
| Vertex AI supported models |
Text, Image, File |
No change |
| Anthropic supported models |
Text, Image, File |
No change |
| Azure AI Foundry supported models |
Text, Image, File |
No change |
- PDF support depends on the model exposing File as a supported input category.
- This update enables PDF processing only for models that already support PDF file inputs through their provider APIs.
- Provider-specific limitations (such as maximum supported PDF size) continue to apply. Some providers enforce maximum file size limits when processing PDF documents. For example, Vertex AI supports up to 15 MB file size.