MetaLanguage model
Code Llama 70B Instruct
Code Llama 70B Instruct has no overall index value yet because there are too few independent benchmark results.
16out of 100-7 vs. last month
based on search, usage and attention
Summary
Code Llama 70B Instruct is an older open Meta checkpoint explicitly tuned for programming tasks. Its Hugging Face card describes downloadable weights under the Llama 2 Community License, not one particular current API service.
The Hugging Face card currently lists no Inference Provider, describes fine-tuning up to 16,000 tokens, and rules out a 100K long-context configuration for this variant. Llama 3.3 70B is a later open Meta checkpoint for text and code, but it remains a different model from Code Llama 70B Instruct.
All benchmark results
Every published result with its source, test setup and comparison models.
Show all 90 published benchmark results
Page 1 of 8
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.
HumanEval (Show measurement, test conditions, and source)
- Source value
- 67.8
- Score
- 67.8
- Metric
- pass@1
- Unit
- %
- Direction
- Higher is better
- Source type
- independent-evaluator
- Evaluator
- Databricks DBRX release benchmark report
- Status
- active
- Retrieved at
- 2026-10-15
- Methodology
- Independent evaluation. Comparison labels the exact CodeLlama-70B-Instruct model; the report text does not provide the evaluation harness, prompt, or decoding parameters for this comparison.
HumanEval (Show measurement, test conditions, and source)
- Source value
- 67.8
- Score
- 67.8
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
HumanEval (Show measurement, test conditions, and source)
- Source value
- 90.3
- Score
- 90.3
- Metric
- pass@10
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; zero-shot; nucleus sampling p=0.95, temperature=0.8.
- Context
- The source does not provide a sample count or confidence interval for this result.
HumanEval (Show measurement, test conditions, and source)
- Source value
- 97.257
- Score
- 97.257
- Metric
- pass@100
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; zero-shot; nucleus sampling p=0.95, temperature=0.8.
- Context
- The source does not provide a sample count or confidence interval for this result.
MBPP (Show measurement, test conditions, and source)
- Source value
- 62.2
- Score
- 62.2
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; 3-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
MBPP (Show measurement, test conditions, and source)
- Source value
- 79.6
- Score
- 79.6
- Metric
- pass@10
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; 3-shot; nucleus sampling p=0.95, temperature=0.8.
- Context
- The source does not provide a sample count or confidence interval for this result.
MBPP (Show measurement, test conditions, and source)
- Source value
- 89.208
- Score
- 89.208
- Metric
- pass@100
- Unit
- %
- Benchmark version
- Code Llama technical report Table 2
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; 3-shot; nucleus sampling p=0.95, temperature=0.8.
- Context
- The source does not provide a sample count or confidence interval for this result.
MultiPL-E HumanEval C++ (Show measurement, test conditions, and source)
- Source value
- 53.4
- Score
- 53.4
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 4
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; C++ translation; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
MultiPL-E HumanEval Java (Show measurement, test conditions, and source)
- Source value
- 58.2
- Score
- 58.2
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 4
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; Java translation; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
MultiPL-E HumanEval PHP (Show measurement, test conditions, and source)
- Source value
- 58.4
- Score
- 58.4
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 4
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; PHP translation; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
MultiPL-E HumanEval TypeScript (Show measurement, test conditions, and source)
- Source value
- 39
- Score
- 39
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 4
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; TypeScript translation; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
MultiPL-E HumanEval C# (Show measurement, test conditions, and source)
- Source value
- 36.7
- Score
- 36.7
- Metric
- pass@1
- Unit
- %
- Benchmark version
- Code Llama technical report Table 4
- Direction
- Higher is better
- Source type
- vendor-reported
- Evaluator
- Code Llama technical report
- Status
- active
- Retrieved at
- 2026-10-16
- Methodology
- Vendor-reported result. CodeLlama-70b-Instruct; C# translation; zero-shot; greedy decoding.
- Context
- The source does not provide a sample count or confidence interval for this result.
Page 1 of 8
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
Code Llama 70B InstructMuse Wonder Full 16Codestral (latest) 16Ornith 1.0 31B 16Solar Mini 16
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 Code Llama 70B Instruct, chosen by model line, developer, size, price and release date.
- Llama 3.1 70B InstructMeta · same developer
- Llama 3.1 70BMeta · same developer
- Llama 3.3 70BMeta · same developer
- Llama-3.3-70B-InstructMeta · same developer
- Llama 3.1 405B InstructMeta · same developer
- Llama-3.2-11B-Vision-InstructMeta · same developer
- Llama 3.1 8B InstructMeta · same developer
- Llama 3.1 8BMeta · same developer
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 Code Llama 70B Instruct
Can I run Code Llama 70B Instruct locally?
Yes, the weights are open.
Stated by Meta
- Open weights
- yes
- License
- Llama 2 Community
- Parameters
- 70B
- active per token
- n/a
Sources and data date
Show all 22 sources
Benchmark sources at a glance
Which independent evaluators tested this model. This is coverage, not a quality ranking.
| Evaluator | Test series | Composite indices | Recorded active measurements |
|---|---|---|---|
| LMArena | 0 | 0 | 1 |
Without vendor claims and historical or disputed results; those stay labelled in the single values. 1 measurements are not yet assigned to a test series.
| Type | Evidence and data date |
|---|---|
| Input | Documented modalities. |
| Output | Documented modalities. |
| Research status | Research date October 15, 2026. 17 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication. |
| Not published | independent-human-eval-setup Note Databricks does not identify the prompt, code execution harness, or sampling configuration for the comparison value. |
| Not published | evaluated N Original wording The leaderboard reports preference-vote totals but no fixed task-case N; Votes are not copied to sampleCount. |
| Unresolved | confidence interval Original wording The Score cell displays ±, but the official policy does not define that notation as a confidence interval; no CI field is populated. |
| Unresolved | score direction Original wording The captured source does not explicitly state higher-is-better or lower-is-better; higherIsBetter remains null. |
| Not published | per-model prompts and generation settings Original wording No exact prompt, decoding settings, or serving configuration is published in the captured row. |
| Additional source | Databricks DBRX release benchmark report (retrieved October 15, 2026) |
| Additional source | Code Llama technical report (retrieved October 16, 2026) |
| Additional source | VHDL-Eval: A Framework for Evaluating Large Language Models in VHDL Code Generation (retrieved October 17, 2026) |
| Additional source | EFFIBENCH (NeurIPS 2024 paper) (retrieved October 17, 2026) |
| Additional source | Arena leaderboard (retrieved October 18, 2026) |
| Additional source | InfiBench paper, NeurIPS 2024 Datasets and Benchmarks Track (retrieved October 20, 2026) |
| Additional source | BigCodeBench official leaderboard (retrieved October 28, 2026)12 |
| Additional source | Model metadata (retrieved October 17, 2026)12 |
| Additional source | Model metadata (retrieved October 18, 2026) |
| Additional source | Model metadata (retrieved October 28, 2026)1234 |
| 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 |