Granite 4.1 8B Benchmark Update
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 103.622 tok/s | HumanEval: 0.218%
View sourceibm-granite
Granite-4.1-8B is a 8B parameter long-context instruct model finetuned from Granite-4.1-8B-Base using a combination of open source instruction datasets with permissive license and internally collected synthetic datasets.
Running this yourself: desktop gpu should be enough.
---
Quality Score
---
Arena ELO
8B
Parameters
131K
Context
Sign in to join the discussion
0
Downloads
0
Likes
May 2026
Released
Benchmarks
19
Open Source
1
Recent launch, pricing, benchmark, and API signals linked to this model or its provider.
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 103.622 tok/s | HumanEval: 0.218%
View sourceQuality: 6.4/100 | Price: $0.063/M tokens | Output: 103.622 tok/s | HumanEval: 0.218%
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 109.198 tok/s | HumanEval: 0.218%
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 109.698 tok/s | HumanEval: 0.218%
View sourceQuality: 6.4/100 | Price: $0.063/M tokens | Output: 116.972 tok/s | HumanEval: 0.218%
View sourceQuality: 6.4/100 | Price: $0.063/M tokens | Output: 113.962 tok/s | HumanEval: 0.218%
View sourceQuality: 6.4/100 | Price: $0.063/M tokens | Output: 109.198 tok/s | HumanEval: 0.218%
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 109.698 tok/s | HumanEval: 0.218%
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 116.972 tok/s | HumanEval: 0.218%
Quality: 6.4/100 | Price: $0.063/M tokens | Output: 113.962 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 120.186 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 116.172 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 104.847 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 101.287 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 113.543 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 113.543 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 102.537 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 126.665 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 110.971 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 110.971 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 122.473 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 123.303 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 126.529 tok/s | HumanEval: 0.218%
Quality: 6.7/100 | Price: $0.063/M tokens | Output: 125.714 tok/s | HumanEval: 0.218%
granite-4.1-8b is now available through local Ollama runtime. 128K context window listed. Granite 4 features improved instruction following (IF) and tool-calling capabilities, making them more effective in enterprise applications.