Head-to-Head Pricing Benchmark
Side-by-side pricing and context window comparison for production model selection.
Default Recommendation (120M input + 60M output)
IBM: Granite 4.0 Micro is lower-cost for the default monthly workload scenario.
Adjust the workload in the calculator below to see a live recommendation for your usage.
| Metric | IBM: Granite 4.0 Micro | Meta: Llama 3.2 1B Instruct |
|---|---|---|
| Developer | ibm-granite | meta-llama |
| Context Window | 131,000 | 60,000 |
| Max Output Tokens | — | — |
| Input Cost / 1M Tokens | $0.0170 | $0.0270 |
| Output Cost / 1M Tokens | $0.1100 | $0.2000 |
| Projected Monthly Cost | $8.64 | $15 |
| Vision | ❌ No | ❌ No |
| Tool Calling | ❌ No | ❌ No |
| Structured Output | ❌ No | ❌ No |
| Reasoning | ❌ No | ❌ No |
| Web Search | ❌ No | ❌ No |
| Tokenizer | Other | Llama3 |
| MMLU Score | N/A | 20.0 |
| GPQA | N/A | 19.6 |
Price History
Current Input / 1M
$0.0170
Current Output / 1M
$0.1100
Price History
Current Input / 1M
$0.0270
Current Output / 1M
$0.2000
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
100.0%
Performance History
Current TPS
0.00
Current Latency
0ms
Uptime
100.0%
Adjust your workload to see projected monthly costs.
IBM: Granite 4.0 Micro
$8.64
per month
Lower costMeta: Llama 3.2 1B Instruct
$15
per month
Live Recommendation
IBM: Granite 4.0 Micro is lower-cost at 120M input + 60M output tokens/month.
Continue evaluation with more “A vs B pricing” decision pages.