IBMGranite 4.2 8B
Best configuration ranks #319 of 610 on the BenchLeader Index at 48.6 ±3.1. Last measured 2 Sept 2026. Released 25 Aug 2026.
- Blended price
- $0.108/M
- $0.060 in · $0.250 out
- Output speed
- 120 tok/s
- First answer
- 17 s
- first token 0.39 s
- Context
- 131k
- Overall index49
- Reasoning45
- Coding49
- Knowledge56
- Human preference45
- Long context48
- Composite45
Benchmark results
One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | default | Source |
|---|---|---|
| LMArena Hard Prompts | 1302#238 | LMArena |
| GPQA Diamond (AA)not in index | 63.1%#340 | Artificial Analysis |
| Humanity's Last Exam (AA)not in index | 9.7%#263 | Artificial Analysis |
Coding
| Benchmark | default | Source |
|---|---|---|
| LMArena Coding | 1364#205 | LMArena |
| SciCode (AA)not in index | 31.5%#145 | Artificial Analysis |
Knowledge
| Benchmark | default | Source |
|---|---|---|
| AA-Omniscience | -17.2#171 | Artificial Analysis |
Human preference
| Benchmark | default | Source |
|---|---|---|
| LMArena Text | 1288#240 | LMArena |
Long context
| Benchmark | default | Source |
|---|---|---|
| AA-LCR | 45.0%#269 | Artificial Analysis |
Composite
| Benchmark | default | Source |
|---|---|---|
| AA Intelligence Index | 12.4#266 | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| CoreWeave | 64 tok/s | 0.20 s | $0.100 | $0.150 | 131k | bf16 |
| DeepInfra | 31 tok/s | 0.57 s | $0.060 | $0.250 | 131k | bf16 |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0001 | – | 19.8 s |
| Summarise a 30-page report | 12,000 / 600 | $0.0009 | – | 22.3 s |
| Code edit | 6,000 / 1,500 | $0.0007 | – | 29.8 s |
| Agentic coding session | 60,000 / 4,000 | $0.0046 | – | 50.7 s |
| Structured extraction | 2,000 / 200 | $0.0002 | – | 18.9 s |
See also
Data as of 9 Sept 2026. Compare with another model.