Granite 4.2 8B
- Context
- 131K
- Listed
- 1mo ago
- Parameters
- 8B
- Input / 1M tokens
- $0.06
- Output / 1M tokens
- $0.25
- Cached input / 1M tokens
- $0.02
Sources:OpenRouter
Benchmarks
DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.
Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.
No benchmark results have been matched to this model yet.
Specifications
- Developer
- IBM
- Model family
- Granite
- Type
- Language model
- Predecessor
- —
- Successor
- —
- Listed on OpenRouter
- August 31, 2026
- Model ID
ibm-granite/granite-4.2-8b- Openness
- Open weights
- License
- —
- Parameters
- 8B
- Input modalities
- text
- Output modalities
- text
- Modality
- text->text
- Native context
- 128K
- Tokenizer
- Other
- Max output tokens
- 118K
- Hugging Face
- ibm-granite/granite-4.2-8b
- Source freshness
- Endpoints synced Oct 15, 2026, 11:24 AM UTC
Open-weight builds
Ollama variants
Every published Ollama tag matched to this model, including parameter editions, instruction/base variants, and quantized builds.
| Variant | Parameters | Type | Quantization | Size | Context | Input |
|---|---|---|---|---|---|---|
| granite4.2:3b-bf16 | 3B | — | BF16 | 7.3 GB | 128K | text |
| granite4.2:3b-q2_K | 3B | — | Q2_K | 1.5 GB | 128K | text |
| granite4.2:3b-q3_K_L | 3B | — | Q3_K_L | 2 GB | 128K | text |
| granite4.2:3b-q3_K_M | 3B | — | Q3_K_M | 1.8 GB | 128K | text |
| granite4.2:3b-q3_K_S | 3B | — | Q3_K_S | 1.7 GB | 128K | text |
| granite4.2:3b-q4_0 | 3B | — | Q4_0 | 2.1 GB | 128K | text |
| granite4.2:3b-q4_1 | 3B | — | Q4_1 | 2.3 GB | 128K | text |
| granite4.2:3b | 3B | — | Q4_K_M | 2.2 GB | 128K | text |
| granite4.2:3b-q4_K_S | 3B | — | Q4_K_S | 2.1 GB | 128K | text |
| granite4.2:3b-q5_0 | 3B | — | Q5_0 | 2.6 GB | 128K | text |
| granite4.2:3b-q5_1 | 3B | — | Q5_1 | 2.8 GB | 128K | text |
| granite4.2:3b-q5_K_M | 3B | — | Q5_K_M | 2.6 GB | 128K | text |
| granite4.2:3b-q5_K_S | 3B | — | Q5_K_S | 2.6 GB | 128K | text |
| granite4.2:3b-q6_K | 3B | — | Q6_K | 3 GB | 128K | text |
| granite4.2:3b-q8_0 | 3B | — | Q8_0 | 3.9 GB | 128K | text |
| granite4.2:8b-bf16 | 8B | — | BF16 | 18 GB | 128K | text |
| granite4.2:8b-q2_K | 8B | — | Q2_K | 3.4 GB | 128K | text |
| granite4.2:8b-q3_K_L | 8B | — | Q3_K_L | 4.7 GB | 128K | text |
| granite4.2:8b-q3_K_M | 8B | — | Q3_K_M | 4.3 GB | 128K | text |
| granite4.2:8b-q3_K_S | 8B | — | Q3_K_S | 3.9 GB | 128K | text |
| granite4.2:8b-q4_0 | 8B | — | Q4_0 | 5.1 GB | 128K | text |
| granite4.2:8b-q4_1 | 8B | — | Q4_1 | 5.6 GB | 128K | text |
| granite4.2:latest | 8B | — | Q4_K_M | 5.3 GB | 128K | text |
| granite4.2:8b-q4_K_S | 8B | — | Q4_K_S | 5.1 GB | 128K | text |
| granite4.2:8b-q5_0 | 8B | — | Q5_0 | 6.1 GB | 128K | text |
| granite4.2:8b-q5_1 | 8B | — | Q5_1 | 6.6 GB | 128K | text |
| granite4.2:8b-q5_K_M | 8B | — | Q5_K_M | 6.3 GB | 128K | text |
| granite4.2:8b-q5_K_S | 8B | — | Q5_K_S | 6.1 GB | 128K | text |
| granite4.2:8b-q6_K | 8B | — | Q6_K | 7.2 GB | 128K | text |
| granite4.2:8b-q8_0 | 8B | — | Q8_0 | 9.3 GB | 128K | text |
| granite4.2:30b-bf16 | 30B | — | BF16 | 45 GB | 128K | text |
| granite4.2:30b-q2_K | 30B | — | Q2_K | 11 GB | 128K | text |
| granite4.2:30b-q3_K_L | 30B | — | Q3_K_L | 15 GB | 128K | text |
| granite4.2:30b-q3_K_M | 30B | — | Q3_K_M | 14 GB | 128K | text |
| granite4.2:30b-q3_K_S | 30B | — | Q3_K_S | 13 GB | 128K | text |
| granite4.2:30b-q4_0 | 30B | — | Q4_0 | 17 GB | 128K | text |
| granite4.2:30b-q4_1 | 30B | — | Q4_1 | 18 GB | 128K | text |
| granite4.2:30b | 30B | — | Q4_K_M | 18 GB | 128K | text |
| granite4.2:30b-q4_K_S | 30B | — | Q4_K_S | 17 GB | 128K | text |
| granite4.2:30b-q5_0 | 30B | — | Q5_0 | 20 GB | 128K | text |
| granite4.2:30b-q5_1 | 30B | — | Q5_1 | 22 GB | 128K | text |
| granite4.2:30b-q5_K_M | 30B | — | Q5_K_M | 21 GB | 128K | text |
| granite4.2:30b-q5_K_S | 30B | — | Q5_K_S | 20 GB | 128K | text |
| granite4.2:30b-q6_K | 30B | — | Q6_K | 24 GB | 128K | text |
| granite4.2:30b-q8_0 | 30B | — | Q8_0 | 31 GB | 128K | text |
Infrastructure
Provider availability
| Provider | Endpoint | Context | Input /M tokens | Output /M tokens | Cached input /M tokens | Cache write /M tokens | Uptime 1d | Data handling |
|---|---|---|---|---|---|---|---|---|
| deepinfra/bf16 | 131K | $0.06 | $0.25 | $0.02 | — | 100.00% | ZDR: YesTraining: Unknown Prompt logging: Unknown Retention: Unknown Checked 2026-10-15 | |
| coreweave/bf16 | 131K | $0.10 | $0.15 | $0.05 | — | 100.00% | ZDR: YesTraining: Unknown Prompt logging: Unknown Retention: Unknown Checked 2026-10-15 |
Provider data:OpenRouter
Deployment graph
Can I run this?
Vertical Tabs only shows compatibility when it has structured evidence tying this exact model to hardware, runtime, quantization, memory use, and observed performance.