ACL 2026: The 64th Annual Meeting of the Association for Computational Linguistics

Registration for ACL 2026 opened on April 21, 2026. The Manchester Grand Hyatt San Diego has a maximum capacity of 3,500 in-person attendees per day. ACL 2025 in Vienna had over 5,000 in-person attendees and the venue was contracted before that growth, so organisers have warned the conference is very likely to sell out quickly. Those planning to attend in person are strongly encouraged to register early. Onsite registration starting July 1 is unlikely to be available for the main conference and may only be available for workshops and tutorials.

The ACL 2026 programme chairs issued a formal statement that papers containing hallucinated references will be desk rejected. This applies across all submission tracks and reflects a broader effort by the NLP community to address the risks of AI-assisted writing in academic submissions. The full statement is in the ACL statement on hallucinated references.

The list of accepted tutorials for ACL 2026 was published on May 2, 2026. Tutorials run on July 2, the day before workshops begin, and cover specialised and emerging research areas in computational linguistics and NLP. See the full ACL 2026 tutorial list.

The list of accepted workshops for ACL 2026 was published on April 9, 2026. Workshops run July 3 and 4 at the Manchester Grand Hyatt San Diego. See the full ACL 2026 workshop list.

A call for shared interest and affinity group events was issued on May 8, 2026, inviting Special Interest Groups, affinity groups and community organisations within the ACL ecosystem to organise events during the conference week. Applications and details are on the shared interest and affinity group events call.

ACL 2026 offers two subsidy programmes to support participation. D&I subsidies offset registration and travel costs for participants from underrepresented backgrounds or low-income regions, with applications due May 15, 2026. Virtual subsidies cover virtual registration costs for eligible participants, with applications due June 6, 2026. Details are on the ACL 2026 calls for papers and participation.

A call for volunteers was issued on April 11, 2026. Student and early-career volunteers support conference operations in exchange for a complimentary registration. Applications are on the ACL 2026 volunteer applications page.

Sessions cover the training, evaluation, alignment, safety, efficiency and deployment of large language models. This is the dominant research strand across the field and touches every other area of the programme, from language understanding and generation to multimodal systems and reasoning.

Research in natural language understanding covers parsing, semantic analysis, coreference resolution, information extraction and question answering. Language generation covers summarisation, data-to-text, dialogue generation, machine translation and controlled text generation.

Sessions address neural machine translation, low-resource and unsupervised translation, multilingual models, cross-lingual transfer and language-specific NLP for under-represented languages. ACL has a strong tradition of multilingual research and the programme consistently covers a wide range of language families.

Research in dialogue systems covers task-oriented dialogue, open-domain conversation, social chatbots, retrieval-augmented generation for dialogue and the evaluation of conversational systems at scale.

The system demonstrations programme showcases working NLP systems, from early-stage research prototypes to mature deployed applications. The industry track covers real-world implementation of NLP systems, lessons learned from production deployment and insights from industry-scale datasets.

The Student Research Workshop provides a dedicated forum for graduate students and early-career researchers to present work-in-progress and receive structured feedback from senior members of the community. The SRW runs alongside the main conference programme.

Tutorials on July 2 provide researchers with deep coverage of specialised and emerging areas in computational linguistics and NLP. Each tutorial is led by subject-matter experts and is designed to be accessible to researchers familiar with the field but new to the specific topic. Up to three instructors per tutorial receive complimentary registration. See the full ACL 2026 tutorial list.

Workshops on July 3 and 4 are co-located conferences focused on specific NLP sub-topics, each with their own paper submissions, accepted papers, invited speakers and programme. Workshops are the primary venue for sub-community activity within the ACL ecosystem and attract researchers from adjacent fields. See the full ACL 2026 workshop list.

The Welcome Reception on Saturday July 4 opens the main conference social programme. Admission to the reception may be added at an additional fee to any conference registration type. The reception provides an informal networking opportunity ahead of the main conference days.

The main conference runs July 5, 6 and 7, covering oral presentations, poster sessions, the system demonstrations programme, the industry track, the Student Research Workshop and panel discussions. A Social Event runs on Monday July 6 and may be added at an additional fee.

ACL 2026 is a hybrid conference. Virtual attendees access all in-person and virtual sessions through the Underline platform, with recordings available for one month after the event. Virtual registration is available as a standalone option and is offered at free or sliding-scale rates for participants in eligible regions under the virtual subsidy programme.

ACL is one of the two most competitive NLP research venues, alongside EMNLP. Acceptance to the main conference represents peer-reviewed validation from the core NLP research community. The tutorials and workshops give researchers early access to emerging areas before they reach the main conference stage.

The industry track and system demonstrations programme give engineering and product teams direct access to the practitioners deploying NLP at scale, with sessions covering production challenges, real-world evaluation, model serving and the gap between research benchmarks and deployed system performance.

The papers accepted to ACL 2026 represent the research that will shape the next generation of language model capabilities, evaluation frameworks and application architectures. For founders building on top of language models, the ACL proceedings are one of the most important technical signals available on where the field is heading.

The intersection of large language models with enterprise NLP use cases, including information extraction, document understanding, question answering and dialogue, is covered across all tracks at ACL. Teams evaluating, fine-tuning or deploying language models will find direct technical reference across the tutorials, workshops and main conference programme.