Inclusionai API: models, pricing, and specs
Compare every Inclusionai model available through Felo API — prices, context windows, and capabilities in one table. Call any of them with the same API key.
- Models
- 1
- chat ×1
- Input per 1M
- $0
- Output per 1M
- $0
- Largest context
- 1.049M
inclusionAI: Ling3.2 Flash is a native multimodal model for long-context visual understanding and multimodal agent tasks.
inclusionAI: Ling3.2 Flash, from Ant Ling, uses a native multimodal architecture that combines vision and language within the model itself. It accepts text, images, and video as input and produces text output, with a1 million token context window. The model has560 billion total parameters, with25 billion parameters active per token.
This model is suited to complex visual understanding, document and frame analysis,3D understanding and interaction, medical multimodal work, and multimodal long-horizon agent tasks. It can also support visual feedback loops in which the model inspects its own renders. Video input is limited to thirty seconds per clip, one clip per request, and thirty-two sampled frames; requests can contain up to forty images, with each image under16384 by784 pixels.
Ling3.2 Flash is positioned above Ling3.1 Flash in Ant Ling's model selection table. Ant reported that it beat Opus5 on ZEROBench and Vision2Web, but no scores were provided. The API has a global rate limit of two requests per second, and synchronous calls have a90-second timeout; streaming is recommended for production use.
Specifications
- Input per 1M
- $0
- Output per 1M
- $0
- Context
- 1.049M
- Max output
- 262.144K
- Protocols
- messages