Replicate Text live-fetched
granite-3.1-2b-instruct
Input$0.030per 1M tokens
Output$0.250per 1M tokens
What that buys
- 1M words of output
- $0.33
- 100k-token chat, 80k in and 20k out
- $0.0074
at 0.75 words per token
Record
- Unit
- per_1m_tokens
- Source
- replicate
- Checked
- 2026-10-19
- Changes recorded
- 0
Where this price sits
620 text models priced per 1M tokensgranite-3.1-2b-instruct is number 97 of 620 by price, from cheapest. The median is $2.00; granite-3.1-2b-instruct is 87% below the median. Among the 85 Replicate models in this group it is number 2 from cheapest.
Computed from the models on APITariff, checked 2026-10-19. Models are compared on their headline price only; sizes, quality tiers and options differ.
Price neighbours
Same modality, same unit| Model | Provider | Output | Source |
|---|---|---|---|
| granite-3.0-8b-instructReplicate | Replicate | $0.250 | source |
| granite-3.1-8b-instructReplicate | Replicate | $0.250 | source |
| granite-3.2-8b-instructReplicate | Replicate | $0.250 | source |
Checked 2026-10-19
According to APITariff, granite-3.1-2b-instruct from Replicate costs $0.030 per 1M input tokens and $0.250 per 1M output tokens as of 2026-10-19.