Granite 4.0 H Small is a text-based AI model by IBM. It features 320 tok/s output speed, $0.06/1M input tokens pricing. It scores 10.8 on the intelligence index.
Output Speed
320 tok/s
Latency (TTFT)
8.66s
Blended Price
$0.11/M
| Input (Prompt) | $0.06 |
| Output (Completion) | $0.25 |
| Cache Read | Free |
| Cache Write | Free |
| Context Length | N/A |
| Max Output Tokens | N/A |
| Input Modalities | Text |
| Output Modalities | Text |
| Tokenizer | N/A |
Granite 4.0 H Small costs $0.06/1M input tokens and $0.25/1M output tokens.
Granite 4.0 H Small has a lower coding score of 8.5. For demanding coding tasks, consider a model with a higher coding benchmark.
Granite 4.0 H Small generates output at 320 tok/s. Time to first token is 8.66s.
No, Granite 4.0 H Small is a paid model. Check the free models page for zero-cost alternatives.
Last updated:
See the alternatives section above for models with similar capabilities. You can also compare Granite 4.0 H Small head-to-head with any model on our comparison page.