IBM › Granite 4.2 8B › Pricing
$0.06 / $0.25per 1M tokens in / out
$0.024per task at thinking?
#221on OpenRouter by tokens · 6.2B / week
Price sheetper 1M tokens
| Input | $0.06 | DeepInfra |
| Output | $0.25 | |
| Context · max output | 128K · 118K | tokens |
Where to buy2 providers · per 1M tokens
| Provider | Input | Output | Context |
|---|---|---|---|
| DeepInfra | $0.06 | $0.25 | — |
| CoreWeave | $0.10 | $0.15 | — |
In IBM's line-upper 1M in / out
| Granite 4.0 Micro | $0.017 / $0.11 | −62% |
| Granite 4.2 3B | $0.03 / $0.12 | −51% |
| Granite 4.0 H Small | $0.05 / $0.22 | −14% |
| Granite 4.2 8B | $0.06 / $0.25 | this model |
| Granite 4.2 30B | $0.16 / $0.65 | 2.6× |
Cheaper for the same workvs Granite 4.2 8B
| GPT-6 Luna medium | -26% | score ×2.7 | $0.017 |
| GPT-5.6 Luna medium | -34% | score ×2.3 | $0.016 |
| Ling 3.0 Flash VL thinkingOpen | -16% | score ×2.2 | $0.06 / $0.18 |
| Ling 3.0 Flash Fin thinkingOpen | -16% | score ×2 | $0.06 / $0.18 |
| MiMo v2.5 thinkingOpen | -19% | score +97% | $0.019 |
Pricing questions
How much does Granite 4.2 8B cost?›
$0.06 per 1M input tokens and $0.25 per 1M output tokens. One Intelligence Index task costs about $0.024 at the default effort.
Is Granite 4.2 8B free to use?›
There is no free API tier, but the weights are open: you can run it on your own hardware for the cost of compute.
What is a cheaper alternative to Granite 4.2 8B?›
GPT-6 Luna at medium effort costs -26% per task with a score ×2.7 on Humanity's Last Exam.