IBMLanguage model
Granite Code 34B
Granite Code 34B has no overall index value yet because there are too few independent benchmark results.
32out of 100-1 vs. last month
based on two data groups, mostly usage
Summary
Granite Code 34B is IBM’s instruction-tuned code checkpoint for programming tasks, logical reasoning, and coding assistants. Its model card describes fine-tuning on permissive instruction data covering code, mathematics, function calling, and general instructions. With 34 billion parameters and an 8,192-token context, it is a focused code specialist rather than a universal long-context chatbot.
The current model card carries an explicit limitation. IBM does not recommend this checkpoint for new projects and points users toward the current Granite mainline. The weights are open under Apache 2.0.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 56 published benchmark results
Page 1 of 5
Measurements without matching peer values
These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.
HumanEvalPack Explain (Python) (Show measurement, test conditions, and source)
- Source value
- 53
- Score
- 53
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Explain (JavaScript) (Show measurement, test conditions, and source)
- Source value
- 45.1
- Score
- 45.1
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Explain (Java) (Show measurement, test conditions, and source)
- Source value
- 50.6
- Score
- 50.6
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Explain (Go) (Show measurement, test conditions, and source)
- Source value
- 36
- Score
- 36
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Explain (C++) (Show measurement, test conditions, and source)
- Source value
- 42.7
- Score
- 42.7
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Explain (Rust) (Show measurement, test conditions, and source)
- Source value
- 23.8
- Score
- 23.8
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (Python) (Show measurement, test conditions, and source)
- Source value
- 54.9
- Score
- 54.9
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (JavaScript) (Show measurement, test conditions, and source)
- Source value
- 47.6
- Score
- 47.6
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (Java) (Show measurement, test conditions, and source)
- Source value
- 55.5
- Score
- 55.5
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (Go) (Show measurement, test conditions, and source)
- Source value
- 51.2
- Score
- 51.2
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (C++) (Show measurement, test conditions, and source)
- Source value
- 47
- Score
- 47
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
HumanEvalPack Fix (Rust) (Show measurement, test conditions, and source)
- Source value
- 45.1
- Score
- 45.1
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- IBM Granite 34B Code Instruct model card
- Status
- active
- Retrieved at
- 2026-10-17
- Methodology
- Vendor-reported result. Exact row in the pinned IBM Granite 34B Code Instruct 8K model card. HumanEvalPack subtask and language are as named in the table. The card marks these metrics with the field veriefied: false; this is not treated as independent verification.
- Context
- IBM marks veriefied: false. No sample count or confidence interval is published. The source does not state a benchmark-suite version.
Page 1 of 5
Popularity compared
How often the model is searched for, used and mentioned, compared with every model of its category.
Popularity index
The 100 most popular models of this category, 0 to 100, higher is more popular
20 of 101 models
As of November 2026 · Source: gradually.ai popularity index
Over time
Weekly values from October 18, 2026 to November 7, 2026
Granite Code 34BAmazon Nova Lite 32KAT-Coder-Pro V2.5 32Qwen2-72B-Instruct 32Claude Haiku 4.5 32
Method changes in the timeline, a jump at these points is no market event:
- 1November 1, 2026: Value only with search or usage data, 30-day attention
What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.
Similar models
More models of the same kind as Granite Code 34B, chosen by model line, developer, size, price and release date.
All specifications
Architecture
Input and output
Watermarking and provenance
AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.
- AI watermarking
- Unknown
- no source
- Watermarking technique
- Not known
- no source
Licence and access
Frequently asked questions about Granite Code 34B
How large is the context window of Granite Code 34B?
8,192 tokens. That is how much text Granite Code 34B reads at once, about 10 A4 pages.
Page count as a rule of thumb from Artificial Analysis
Can I run Granite Code 34B locally?
Yes, the weights are open.
Stated by IBM
- Open weights
- yes
- License
- Apache 2.0
- Parameters
- 34B
- active per token
- n/a
Sources and data date
Show all 12 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
No active result from the reviewed external evaluators can currently be attributed to this model. Other source evidence remains available below.
Without vendor claims and historical or disputed results; those stay labelled in the single values.
| Type | Evidence and data date |
|---|---|
| Input | Documented modalities. |
| Output | Documented modalities. |
| Research status | Research date October 15, 2026. 5 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | pass-at-one-unit-and-evaluation-parameters Note The card's model-index values lack explicit score units, prompt details, sampling parameters, and run dates; raw values are preserved. |
| Not published | confidence-interval Original wording The paper publishes no confidence intervals for these table cells. |
| Unresolved | checkpoint-revision Original wording The paper maps the evaluated Granite model name to the exact Hugging Face repository but does not bind the evaluation to an immutable repository revision or weight hash. |
| Additional source | IBM Granite 34B Code Instruct model card (retrieved October 17, 2026) |
| Additional source | Lin et al., CodeReviewQA, ACL Anthology (retrieved October 17, 2026) |
| Additional source | IBM Granite Code technical report (arXiv:2405.04324) (retrieved October 23, 2026) |
| Additional source | Granite Code Models: A Family of Open Foundation Models for Code Intelligence (retrieved October 23, 2026) |
| Additional source | Granite Code paper · Editorial description reviewed October 2, 2026 |
| Additional source | OpenRouter: Popularity index (derived), CC BY 4.0 |
| Additional source | Vercel AI Gateway: Popularity index (derived), CC BY 4.0 |
| Additional source | Ollama: cumulative downloads for the popularity index |
| Additional source | Replicate: cumulative runs for the popularity index |
| Additional source | Google Trending Now: limited feed selection for Germany and the US |
| Additional source | Creative Commons Attribution 4.0 |