---
id: CAP-OPEN-001
title: "Open-Weight versus Closed Models"
alternative_names:
  - "Open Foundation Models"
  - "Open-Source AI"
  - "Model Release Strategies"
category: "Capabilities → Landscape"
volume: IV
chapter: 25
entry_type: "Standard Article"
level: 3
status: "Draft"
last_reviewed: 2026-10-18
version: "0.1.0"
prerequisites:
  - "TECH-LLM-001"
related_entries:
  - "CAP-COMPARE-001"
  - "CAP-GENAI-001"
  - "GOV-LAW-001"
  - "SAFE-SEC-001"
  - "SAFE-ALIGN-001"
  - "ECON-COMP-001"
---

# Open-Weight versus Closed Models

## 1. Concise Definition
An **open-weight** model is one whose trained parameters (weights) are published for download, usually under a licence, so that others can run, fine-tune and inspect it locally. A **closed** model is available only through a service interface (an API or product) controlled by its developer. "Open-weight" is narrower than "open-source": the Open Source Initiative's Open Source AI Definition 1.0 requires, in addition to parameters, sufficiently detailed information about the training data and the complete training and inference code (Open Source Initiative, 2024).

## 2. Overview & Significance
Release strategy determines who can use a model, modify it, audit it and misuse it. The choice therefore bears on research reproducibility, market competition, security and regulation. Public debate often treats "open versus closed" as binary, but release is a spectrum, and the terms are used loosely (and sometimes strategically). This article defines the terms, summarises the main arguments and evidence, and notes where law and policy treat the categories differently. It does not rank current models; for dated model facts, see `MODEL-001`.

## 3. Historical Development
Machine learning research has long released code and trained models. In language modelling, a period in which leading developers withheld weights of their largest models while describing them in papers was followed by releases of weights such as LLaMA (Touvron et al., 2023a) and Llama 2 (Touvron et al., 2023b), the latter described by its authors as open foundation and fine-tuned chat models. Other developers released models under permissive licences, for example Mistral 7B (Jiang et al., 2023), and fully open efforts such as OLMo aimed to release not only weights but also data, code and evaluation tooling (Groeneveld et al., 2024). DeepSeek-AI released the DeepSeek-V3 technical report (DeepSeek-AI, 2024) and the R1 reasoning model, whose report was published in *Nature* in 2025 (Guo et al., 2025), demonstrating that openly described reasoning-model training was feasible outside the largest closed developers.

Evaluation of openness itself developed in parallel. The Foundation Model Transparency Index scored developers on disclosure of upstream resources, model properties and downstream use (Bommasani et al., 2023). Liesenfeld and Dingemanse (2024) assessed claims of openness in generative AI and argued that "open washing" exists, where partial releases are marketed as open source, with implications for the EU AI Act. The Open Source Initiative published its definition in 2024.

## 4. Technical / Conceptual Explanation

### 4.1 A spectrum of release
Seger et al. (2023) describe a gradient of release options from fully closed to fully open, including staged release, hosted access, structured API access with safeguards and full open release, and discuss alternative ways of pursuing the aims of open-sourcing without releasing weights. The relevant artefacts are separable:

| Artefact | Closed (API) | Open-weight | Open-source (OSI definition) |
| :--- | :--- | :--- | :--- |
| Hosted inference | Yes | Optional | Optional |
| Weights | Not released | Released | Released under OSI-approved terms |
| Architecture details | Often partial or undisclosed | Usually released | Released |
| Training code | Not released | Varies | Required (complete code) |
| Training data | Not released | Usually not released | Detailed data information required |
| Licence | Terms of service | Custom or permissive licence | OSI-approved terms |

The table is a summary of the definitions cited, not a claim about any particular model. The OSI definition lists three components of the preferred form for modification: data information, code and parameters (Open Source Initiative, 2024).

### 4.2 What weight access enables
Possessing weights allows local inference without data leaving the user's infrastructure, fine-tuning for specialised tasks, quantisation and compression, inspection of internal activations for research, and long-term availability independent of the developer's service. Kapoor et al. (2024) identify five distinctive properties of open foundation models, including greater customisability and poor monitoring, and argue that these properties lead to both benefits and risks.

### 4.3 What closed access enables
A closed service allows the developer to apply safeguards, monitor misuse, update the model, withdraw access and charge per use. The cost is dependence: users cannot audit the weights, are subject to changes and deprecations, and must send their data to the provider.

## 5. Types, Components & Classifications
- **Closed / API-only**: weights withheld; interaction through a service.
- **Open-weight, restrictive licence**: weights downloadable but with conditions on use or scale, which fails the Open Source Definition's freedom to use for any purpose.
- **Open-weight, permissive licence**: weights under a licence such as Apache 2.0 (Jiang et al., 2023, state that Mistral 7B is released under Apache 2.0).
- **Open-source AI (OSI)**: weights, code and data information under OSI-approved terms (Open Source Initiative, 2024).
- **Hybrid**: smaller models open, largest closed; or weights released after a delay.

Classification of a specific model into these categories requires reading its licence and release materials. The licence of a given model version is a fact to be checked at the source, not inferred from marketing.

## 6. Applications & Concrete Examples
Open-weight models are used for on-premises deployment where data cannot leave an organisation, for research on interpretability and training methods, for fine-tuned domain models and for low-cost inference. Closed models are widely used through APIs and consumer products where operators prefer a managed service and developer-provided safeguards. A single organisation may use both. Open-weight releases such as Llama 2, Mistral 7B and OLMo are cited here as representative examples of a type, not as statements about current capability or licensing, which change between versions.

## 7. Advantages, Limitations & Unresolved Questions

**Safety and misuse.** Qi et al. (2023) showed that fine-tuning aligned language models on a small number of adversarial examples can remove safety behaviour, and that even benign fine-tuning can degrade it. This was demonstrated for both an open-weight model (Llama-2-7b-chat) and a closed model offered with a fine-tuning API (GPT-3.5 Turbo), so it does not by itself separate the two cases. For open weights, however, the developer cannot recall a release or enforce usage policies technically once weights are downloaded. Seger et al. (2023) assess risks and benefits of releasing highly capable models and discuss alternatives. Kapoor et al. (2024) argue that the evidence on the *marginal* risk of open models, compared with existing technology and closed models, should be assessed per risk vector, and that much concern lacks empirical backing; they present this as a position paper. Both sides cite the same missing evidence: systematic, replicated studies of real-world misuse uplift are few. The question is therefore **Contested**.

**Competition and concentration.** Open weights lower barriers to entry and allow independent audit; closed models allow developers to recoup training cost and apply safeguards. The cost of frontier training is covered in [ECON-COMP-001]; the sources cited here do not quantify the net effect on market concentration.

**Evaluation fairness.** Comparisons between open and closed models are affected by access asymmetries; for example, Singh et al. (2025) reported that in Chatbot Arena in their study, an estimated 29.7% of data went to a combined 83 open-weight models, against 19.2% and 20.4% for two large proprietary providers each (2025 estimates; see [CAP-COMPARE-001]).

**Terminology.** Calling a model "open source" when training data or code are withheld contradicts the OSI definition; Liesenfeld and Dingemanse (2024) discuss this as open washing.

## 8. Current State of Knowledge
As of 2026-10-15: multiple developers have released open-weight models of significant capability, with a range of licences and disclosure levels (Well Supported; Touvron et al., 2023b; Jiang et al., 2023; Groeneveld et al., 2024; Guo et al., 2025). A formal definition of open-source AI exists and is stricter than weight release (Established as a published definition; adoption is voluntary). The magnitude of any capability gap between the best open-weight and best closed models varies with date, domain and measurement; this article gives no figure and defers to `MODEL-001` and the cautions in [CAP-COMPARE-001]. The balance of risks and benefits of open weights for highly capable models is Contested.

**Law and policy (checked against the Act's text as published on a mirror of the Regulation, see references).** Regulation (EU) 2024/1689 (the EU AI Act) states in Article 2(12) that it does not apply to AI systems released under free and open-source licences unless they are placed on the market or put into service as high-risk systems or are covered by certain other provisions. Article 53(2) exempts providers of general-purpose AI models released under a free and open-source licence, with publicly available parameters, architecture information and usage information, from some documentation obligations, but this exception does not apply to general-purpose AI models with systemic risk. Chapter V on general-purpose AI models applies from 2 August 2025 under Article 113(b) as shown at the cited source; the Act has been subject to amendment proposals and the current application dates should be checked at the official EU source [VERIFICATION NEEDED: current in-force text and dates of Articles 2, 53 and 113 against EUR-Lex]. On United States policy, Executive Order 14110 was revoked on 2025-01-20 [VERIFICATION NEEDED: primary source for revocation; stated in the project brief]. See [GOV-LAW-001].

## 9. Future Trajectory & Strategic Outlook
Observed pattern, not a forecast: release strategies are mixed, with developers releasing some models open and keeping others closed, and with the definitions of "open" contested (Open Source Initiative, 2024; Liesenfeld & Dingemanse, 2024). The direction of regulation, which treats openness as a partial exemption but not for systemic-risk models in the EU, is documented above; whether other jurisdictions adopt comparable carve-outs is open. This article registers no forecast.

## 10. Related Entries & Knowledge Graph
- **Foundational Concepts**:
  - [[TECH-LLM-001] Large Language Models](../vol03_technical/TECH-LLM-001_large_language_models.md): The model class to which release choices apply.
  - [[CAP-GENAI-001] Generative AI](CAP-GENAI-001_generative_ai.md): Open and closed generative systems across modalities.
- **Related Methods & Technologies**:
  - [[CAP-COMPARE-001] Methodology for Comparing AI Systems](CAP-COMPARE-001_methodology_for_comparing_ai_systems.md): Fair comparison across release types.
  - [[ECON-COMP-001] Compute Economics](../vol09_economy/ECON-COMP-001_compute_economics.md): Training cost and release incentives.
  - [[SAFE-SEC-001] AI Security: Prompt Injection, Poisoning, Extraction](../vol08_safety_governance/SAFE-SEC-001_ai_security_prompt_injection_poisoning_extraction.md): Security consequences of weight access.
- **Advanced Implications & Future Trajectories**:
  - [[GOV-LAW-001] AI Governance](../vol08_safety_governance/GOV-LAW-001_ai_governance.md): Legal treatment of open models.
  - [[SAFE-ALIGN-001] The AI Alignment Problem](../vol08_safety_governance/SAFE-ALIGN-001_ai_alignment.md): Safety behaviour persistence after release.
  - [[MODEL-001] Model Registry](../reference/factbook/MODEL-001_model_registry.md): Dated model facts and licences.

## 11. References & Foundational Literature
1. Bommasani, R., Klyman, K., Longpre, S., Kapoor, S., et al. (2023). *The Foundation Model Transparency Index*. arXiv:2310.12941. [Scores foundation-model developers on transparency].
2. DeepSeek-AI. (2024). *DeepSeek-V3 Technical Report*. arXiv:2412.19437. [Technical report accompanying an open-weight model release].
3. European Union. (2024). *Regulation (EU) 2024/1689 laying down harmonised rules on artificial intelligence (Artificial Intelligence Act)*. Official Journal of the EU. http://data.europa.eu/eli/reg/2024/1689/oj (Official Journal text of Articles 2(12) and 53(2) read 2026-10-18; Article 113 as adopted, later amended by Regulation (EU) 2026/1744). [Articles 2(12), 53(2) and 113 as cited in Section 8].
4. Groeneveld, D., Beltagy, I., Walsh, P., Bhagia, A., et al. (2024). *OLMo: Accelerating the Science of Language Models*. arXiv:2402.00838. [Open language model release oriented to open research].
5. Guo, D., et al. (DeepSeek-AI). (2025). *DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning*. Nature, 645, 633–638. https://doi.org/10.1038/s41586-025-09422-z. [Peer-reviewed report on an open reasoning model].
6. Jiang, A. Q., Sablayrolles, A., Mensch, A., Bamford, C., et al. (2023). *Mistral 7B*. arXiv:2310.06825. [Open model release (Apache 2.0 licence stated in the report)].
7. Kapoor, S., Bommasani, R., Klyman, K., Longpre, S., et al. (2024). *On the Societal Impact of Open Foundation Models*. arXiv:2403.07918. [Position paper on benefits and risks of widely available weights].
8. Liesenfeld, A., & Dingemanse, M. (2024). *Rethinking open source generative AI: open washing and the EU AI Act*. Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency, 1774–1787. https://doi.org/10.1145/3630106.3659005. [Assessment of claims of openness and relation to the EU AI Act].
9. Open Source Initiative. (2024). *The Open Source AI Definition 1.0*. https://opensource.org/ai/open-source-ai-definition. [Defines requirements for data information, code and parameters].
10. Qi, X., Zeng, Y., Xie, T., Chen, P.-Y., et al. (2023). *Fine-tuning Aligned Language Models Compromises Safety, Even When Users Do Not Intend To!* arXiv:2310.03693. [Fine-tuning can degrade safety alignment in open and API-fine-tunable models].
11. Seger, E., Dreksler, N., Moulange, R., Dardaman, E., et al. (2023). *Open-Sourcing Highly Capable Foundation Models: An evaluation of risks, benefits, and alternative methods for pursuing open-source objectives*. arXiv:2311.09227. [Risk-benefit evaluation and alternatives to full open release].
12. Singh, S., Nan, Y., Wang, A., D'Souza, D., et al. (2025). *The Leaderboard Illusion*. arXiv:2504.20879. [Estimated data-share asymmetries by provider type in Chatbot Arena].
13. Touvron, H., Lavril, T., Izacard, G., Martinet, X., et al. (2023a). *LLaMA: Open and Efficient Foundation Language Models*. arXiv:2302.13971. [Early weight release of a foundation language model family].
14. Touvron, H., Martin, L., Stone, K., Albert, P., et al. (2023b). *Llama 2: Open Foundation and Fine-Tuned Chat Models*. arXiv:2307.09288. [Weights released with a custom licence].

## 12. Evidential Classification Matrix
| Section / Claim | Evidence Category | Supporting Source | Confidence / Status |
| :--- | :--- | :--- | :--- |
| Open-weight is narrower than OSI-defined open-source AI | Established (published definition) | Open Source Initiative (2024) | High |
| Release is a spectrum with several intermediate options | Well Supported (conceptual analysis) | Seger et al. (2023) | Moderate to High |
| Fine-tuning can remove safety behaviour in open and API-tunable models | Well Supported (reproduced across two model types in one study) | Qi et al. (2023) | Moderate |
| Marginal risk of open models is low or unproven | Contested | Kapoor et al. (2024); Seger et al. (2023) | Active debate |
| Arena data-share asymmetries between open and proprietary models | Preliminary (single 2025 study) | Singh et al. (2025) | Moderate |
| EU AI Act Articles 2(12) and 53(2) open-source provisions | Established (statutory text); dates need re-check | Regulation (EU) 2024/1689 | High for text; Moderate for dates |
| EO 14110 revoked 2025-01-20 | Established per project brief; primary source not checked | [VERIFICATION NEEDED] | Moderate |

## 13. Revision History
- **v0.1.0** (2026-10-15): Initial draft. Citations checked against arXiv/Crossref/publisher pages where marked; not independently fact-checked.