Massive stars are the engines of the Cosmos, shaping their environments and driving galaxy evolution across cosmic time. Yet, this general textbook picture faces many challenges when trying to turn abstract insights into quantitative predictions. Recent discoveries, such the surprisingly high metallicity and early nitrogen enrichment in high-redshift galaxies discovered by JWST, are challenging current descriptions of massive star evolution and add new pieces to a puzzle that is yet everything but complete.

The oncoming era of large surveys and advances in computational modeling create the potential to reach breakthroughs in our understanding. Yet, to resolve current problems and conflicting conclusions, we will also need to reconsider what we think we know. Are the objects we observe what we think they are? Are the models we use describing what is actually going on? And what can we learn from previous misconceptions? This short review highlights major open questions from individual stars and stellar systems back to the first galaxies while also discussing two examples – the weak-wind problem as well as the different flavours and impact of Wolf-Rayet stars – where recent discoveries might point in a new direction.

We present an overview of the main characteristics of several spectroscopic surveys designed to advance our understanding of the physical properties and evolution of massive stars. We also summarize key results obtained from the analysis of these datasets, highlighting how the interpretation of some observables in the framework of massive-star evolution is considerably more complex than previously anticipated.

I present some of the highlights of the Gaia mission on massive stars and discuss what the fourth data release (DR4) will bring in late 2026. In the first part of the contribution I describe the different types of Gaia products available now and for DR4 and their caveats. In the second part I present the most significant results on massive stars regarding parameter determination, binaries, photometric variability, stellar groups, and the sample and spatial distribution.

XShootU is an open world-wide collaboration aiming to provide high-quality optical to near-infrared spectra for all massive stars in ULLYES. Here, an overview is given of the available and upcoming data releases, as well as the scientific output so far.

We present a new spectroscopic pipeline designed to analyse large numbers of hot massive stars homogeneously. The pipeline has been developed to utilise large grids of FASTWIND non-LTE, line blanketed models in which spherical geometry is adopted, and uniquely incorporates model errors. The pipeline has been applied to three contemporary datasets involving Very Large Telescope spectroscopy of OB stars in the Magellanic Clouds, namely the VLT FLAMES Tarantula Survey (VFTS), XShooting-ULLYSES (XShootU) and Binaries at Low Metallicity (BLOeM). We find satisfactory agreement with previous detailed temperatures and surface gravities, although strong nebular contamination, binarity and disk emission from OBe stars are problematic for automatic pipelines, requiring visual inspection of fits. The tool has been incorporated into the pipeline for the VISTA/4MOST pipeline.

We present an overview of our recent results from the BLOeM campaign in the Small Magellanic Cloud ($Z=0.2\,{\rm Z}_\odot$). Using nine-epoch VLT/FLAMES spectroscopy, we investigated the multiplicity of 929 massive stars. Our findings reveal contrasting binary properties across evolutionary stages: O-type stars show an intrinsic close-binary fraction of 70%, and early B-type dwarfs/giants reach $\sim80\%$, exceeding higher-metallicity samples. In contrast, B0–B3 supergiants drop to $\sim40\%$, and A–F supergiants to $\sim8\%$; intrinsic variability likely inflates the latter, so the true multiplicity may be lower. OBe stars display distinct binary properties consistent with a post-interaction origin. These results have profound implications for massive-star evolution at low metallicity, including the production of exotic transients, gravitational-wave progenitors, and ionising radiation in the early Universe.

In the era of large spectroscopic surveys, a vast amount of spectra of massive stars will be gathered and supplemented by the wealth of astrometric and photometric data provided by the Gaia satellite. Released data will mean a major step forward in the study of massive stars, giving us the chance to create statistically significant samples to explore the role of almost any parameter. In this contribution, I introduce to the community the Multi-wavelength Exploration of massIve star-forminG regions and ASsociations project (MEIGAS) and long-term plans for conducting comprehensive studies in the major galactic and near extragalactic star-forming regions and OB associations. Benefiting from current and forthcoming data from large scale spectroscopic surveys such as WEAVE and 4MOST (among others), as well as complementary observations at different wavelength ranges, the project aims to achieve crucial and complementary information to adequately characterize these regions and their stellar content, something imperative to improve our understanding of star formation and poorly known evolutionary pathways of massive stars.

Massive stars are pivotal drivers of galactic chemo-dynamical evolution. Despite their relevance, several key aspects of their nature and evolution remain poorly constrained. Fortunately, large-scale spectroscopic surveys combined with Gaia kinematics offer the opportunity to analyze the physical, chemical, kinematical, and binary properties of hundreds of massive stars in the Milky Way. However, studying mixed populations born under different initial conditions can lead to misinterpretation of the results. In this context, young open clusters and associations are ideal laboratories given the similar age and initial composition of their stellar populations. One of the known young associations with more massive stars in the Milky Way is Per OB1. In this work, we present recent findings from the quantitative spectroscopic analysis of a thousand Galactic OB stars observed within the IACOB project and compare some of these findings with the properties of Per OB1’s complete population of luminous massive stars.

The Milky Way Mapper (MWM) in SDSS-V includes a massive star program (OBA spectral types) expected to obtain multi-epoch optical and near-infrared spectra for $\sim10^5$ OBA targets in the Milky Way and the Magellanic Clouds. This program will offer an unprecedented panorama of their properties across the Galactic disk ecosystem, allowing us to detail kinematic distributions, their impact on the interstellar medium (ISM), and their relation to the evolution of young star clusters. The SDSS consortium is also exploring the interstellar medium in the Milky Way, the Magellanic Clouds, and galaxies of the Local Group with the Local Volume Mapper (LVM) program. The LVM is mapping the warm ionized ISM through 3D spectroscopy (3600–9800 Å, R$\sim$4000) using a state-of-the-art wide field integral spectroscopy unit (IFU; 0.165 sq deg, 1800 fibers), aiming to explore hundreds of H II regions, from the solar neighborhood to nearby galaxies. The MWM-LVM synergy sets one magnificent stage for OB star science in the next few years.

Understanding the processes in place during the epoch of reionization (EoR) is a frontier goal of observational cosmology, centered on uncovering the sources of reionization. Local reionization-era analogs are commonly used as proxies for EoR galaxies; however, these analogs are often too massive. The low-mass dwarf starburst galaxies Pox 186 and UM 461 offer physical conditions representative of EoR systems. We discuss comprehensive panchromatic data from the far-UV to the mid-IR, emphasizing ground-based IFU, HST and JWST observations. We model the stellar population, nebular emission, and dust using various stellar and nebular synthesis codes, and address challenges in these models to inform the next generation of stellar atmosphere and evolution models.

We present the properties of six Type II supernovae ($0.675 \leq z \leq 3.61$) discovered by the JWST Advanced Deep Extragalactic Survey (JADES) transient survey through light-curve modeling. Two SNe exhibit high explosion energies (around $3\times10^{51}$ erg), while four have typical energies ($(0.5-2)\times10^{51}$ erg). The fraction of high-energy Type II SNe may be higher at high redshifts due to lower metallicity, but small sample size and observational biases limit firm conclusions. We found that introducing confined, dense circumstellar matter improves light-curve fits in some cases, suggesting its frequent presence at high redshifts. More high-redshift Type II SNe are required to explore differences between nearby and distant events, but this work offers the first glimpse into the high-redshift population.

A robust stellar initial mass function (IMF) is crucial in any studies related to star formation. However, the direct measurement of the stellar IMF is confined to the local universe, limited by the resolving power of telescopes. Recently, a new method for accessing the stellar IMF beyond the local universe has been developed. The observed detection rate of transient lensed stars – individual, massive, thus luminous stars in strongly lensed galaxies that are temporarily detectable upon stellar microlensing – can serve as a probe to break the IMF-star formation history degeneracy in studies utilizing spectral energy distribution fitting, hence providing a window to look at the IMF at a subsample of gravitationally lensed galaxies. In this proceeding, I summarize the contributed talk given at IAUS402 entitled the same as this contribution and highlight some key results, which currently show no evidence for a top-heavy IMF in $z \approx 1$ galaxies.

Born with masses greater than 100 $M_\odot$, very massive stars (VMS) are extreme sources of radiative, mechanical, and chemical feedback. Due to their high luminosity, individual VMS can dominate the integrated light of stellar populations, emphasizing the need for accurate models of such stars, particularly in the context of JWST observations of unresolved galaxies in the early universe. Here we present our recent work supplementing evolutionary models computed in GENEC with detailed atmosphere models computed in PoWR to obtain a more comprehensive picture of VMS. As part of this effort, we have computed a series of deep atmospheric models of VMS that cover the entire region of the hot iron bump. Our results show that this approach avoids the opacity-induced inflation predicted by the evolutionary code during the main sequence, which significantly impacts the predicted effective temperatures of VMS and reveals new uncertainties in spectroscopic diagnostics.

Unveiling massive stars’ internal structure and the physical origin and efficiency of the internal mixing processes? It is now possible using the apsidal motion rate in close eccentric binaries! The apsidal motion rate depends on the tidal interactions occurring between the stars and is proportional to $k_2$, a measure of the star’s inner density profile. Confronting standard stellar models with observations reveals the famous $k_2$-discrepancy: models predict too high a $k_2$ for the stars, that is to say, stars with too low a density contrast between their core and envelope. We built bespoke GENEC stellar evolution models including tidal mixing for the twin massive binary HD 152248. The models reveal the instabilities allowing to reproduce the stellar density profiles: advecto-diffusive models better reproduce $k_2$ than magnetic models. A large overshooting is necessary to converge towards the observed $k_2$, yet alone is not sufficient. While a change in metallicity or mass-loss rate has no significant impact on $k_2$, a larger initial helium abundance allows us to better reproduce the $k_2$. Yet, a super-solar helium abundance is not observationally supported. Our analyses highlight the need for a process in the stars that slows down the radial expansion.

Realistic model atmospheres have progressively been included through various methods in state-of-the-art low- and intermediate-mass stars evolution calculations. At the same time model atmospheres for massive stars have become increasingly realistic over the last 50 years, with the progressive inclusion of non-LTE and non-gray effects, wind extension, and line-blanketing. However all these improvements have never been included in massive star evolution calculation, which in their vast majority still use the Eddington gray approximation. We describe here a method to include realistic model atmospheres in the evolution calculations of massive stars. We present preliminary results of the application of this method to a $60 M_{\odot}$ star at solar metallicity. These results show that for these mass and metallicity the Eddington gray approximation is valid for the main-sequence evolution.

Massive star evolution in binaries frequently leads to the formation of stripped stars that have lost a significant part of their hydrogen-rich envelope after mass transfer via Roche-lobe overflow. This paper summarizes the observational and modeling progress made in identifying and characterizing a unique class of post-interaction objects: partially stripped stars orbiting Be companions. We highlight how multi-wavelength and multi-epoch spectroscopy, combined with non-LTE atmosphere modeling, is essential for understanding these binary systems. Key results demonstrate that these objects are systematically over-luminous for their low masses, exhibit surface CNO-processed material, and drive exceptionally powerful stellar winds. The recent analysis of the Galactic Be binary MWC 656, revealing a hot stripped star instead of a black hole companion, is presented as a case study. The growing sample provides vital empirical constraints to refine binary evolution models, particularly concerning wind mass-loss, mass transfer efficiency, and the formation of compact objects.

Model atmosphere analyses provide observational constraints that can be used to test models of OB stars to answer key questions of massive star research. In the focus are elemental abundances and atmospheric parameters, which in particular in the era of precise and accurate parallax measurements by the Gaia mission can be used to derive fundamental stellar parameters. An overview of the state-of-the-art of quantitative spectroscopy of OB stars is given, concentrating on one hand on the question of how well information can be extracted from observations, which is of particular interest on the eve of the upcoming large-scale spectroscopic surveys. On the other hand, existing observational results are confronted with model predictions, illuminating successes but also pointing out the open questions.

Mass loss is a key mechanism affecting the evolution of massive stars. However, the real mass-loss rates in the low-metallicity, low-luminosity regime have yet to be constrained and mostly have upper limits derived from often insensitive or degenerate diagnostics. The derived wind momenta are discrepant to the theoretical predictions that are a factor of 1 to 2 orders of magnitude higher.

We present the observations of a sample of 15 SMC O-stars observed with JWST/NIRSpec, and the preliminary spectroscopic results from a subsample of 5 stars. The full sample spans from the weak to strong-wind regimes at SMC metallicity (20% solar), with an emphasis on the Br$\alpha$ spectral line as a sensitive mass-loss diagnostic for weak-winded stars. The observation of this near infrared mass-loss diagnostic is impossible in SMC stars from ground-based facilities, but is now enabled by JWST’s superb sensitivity.

We analyzed 5 stars within our full sample and found consistent mass loss rates between the CMFGEN fits, FASTWIND fits, and UV/optical literature values. These preliminary results confirm the position of these 5 stars on the observed WLR at SMC metallicity, thus supporting that the weak-wind discrepancy is real in the thin-wind regime, as well as demonstrating the robustness of Br$\alpha$ as a mass-loss diagnostic for massive stars.

The B[e] phenomenon discovered nearly 50 years ago features the presence of forbidden emission lines due to extended and dense circumstellar gas and large IR excesses due to the radiation from circumstellar dust in a wide variety of objects from pre-main-sequence stars to Planetary Nebulae. It also shows up in a small group of supergiants that includes Luminous Blue Variables, such as $\eta$ Carinae. Over the years, some of them were proven to be binary systems, but the presence of a secondary component in other is still elusive. At the same time, there is growing evidence that the B[e] phenomenon can be due to binary mergers or interactions in triple systems.